Skip to content

P
PaddleDetection

PaddlePaddle / PaddleDetection
大约 1 年 前同步成功

通知 695
Star 11112
Fork 2696
P
PaddleDetection
未验证 提交 bf79506a 编写于 作者: W wangxinxin08 提交者： GitHub
浏览文件
操作

modify links about master, test=document_fix (#3318)

上级 c2bc4b4a
展开全部 隐藏空白更改
内联 并排
Showing with 319 addition and 319 deletion
configs/ppyolo/README_cn.md
浏览文件 @ bf79506a
...@@ -55,7 +55,7 @@ PP-YOLO从如下方面优化和提升YOLOv3模型的精度和速度： ...@@ -55,7 +55,7 @@ PP-YOLO从如下方面优化和提升YOLOv3模型的精度和速度：
**注意:** **注意:**
- PP-YOLO模型使用COCO数据集中train2017作为训练集，使用val2017和test-dev2017作为测试集，Box AP<sup>test</sup>为`mAP(IoU=0.5:0.95)`评估结果。 - PP-YOLO模型使用COCO数据集中train2017作为训练集，使用val2017和test-dev2017作为测试集，Box AP<sup>test</sup>为`mAP(IoU=0.5:0.95)`评估结果。
- PP-YOLO模型训练过程中使用8 GPUs，每GPU batch size为24进行训练，如训练GPU数和batch size不使用上述配置，须参考[FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/FAQ.md)调整学习率和迭代次数。 - PP-YOLO模型训练过程中使用8 GPUs，每GPU batch size为24进行训练，如训练GPU数和batch size不使用上述配置，须参考[FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/docs/tutorials/FAQ.md)调整学习率和迭代次数。
- PP-YOLO模型推理速度测试采用单卡V100，batch size=1进行测试，使用CUDA 10.2, CUDNN 7.5.1，TensorRT推理速度测试使用TensorRT 5.1.2.2。 - PP-YOLO模型推理速度测试采用单卡V100，batch size=1进行测试，使用CUDA 10.2, CUDNN 7.5.1，TensorRT推理速度测试使用TensorRT 5.1.2.2。
- PP-YOLO模型FP32的推理速度测试数据为使用`tools/export_model.py`脚本导出模型后，使用`deploy/python/infer.py`脚本中的`--run_benchnark`参数使用Paddle预测库进行推理速度benchmark测试结果, 且测试的均为不包含数据预处理和模型输出后处理(NMS)的数据(与[YOLOv4(AlexyAB)](https://github.com/AlexeyAB/darknet)测试方法一致)。 - PP-YOLO模型FP32的推理速度测试数据为使用`tools/export_model.py`脚本导出模型后，使用`deploy/python/infer.py`脚本中的`--run_benchnark`参数使用Paddle预测库进行推理速度benchmark测试结果, 且测试的均为不包含数据预处理和模型输出后处理(NMS)的数据(与[YOLOv4(AlexyAB)](https://github.com/AlexeyAB/darknet)测试方法一致)。
- TensorRT FP16的速度测试相比于FP32去除了`yolo_box`(bbox解码)部分耗时，即不包含数据预处理，bbox解码和NMS(与[YOLOv4(AlexyAB)](https://github.com/AlexeyAB/darknet)测试方法一致)。 - TensorRT FP16的速度测试相比于FP32去除了`yolo_box`(bbox解码)部分耗时，即不包含数据预处理，bbox解码和NMS(与[YOLOv4(AlexyAB)](https://github.com/AlexeyAB/darknet)测试方法一致)。
...@@ -68,7 +68,7 @@ PP-YOLO从如下方面优化和提升YOLOv3模型的精度和速度： ...@@ -68,7 +68,7 @@ PP-YOLO从如下方面优化和提升YOLOv3模型的精度和速度：
| PP-YOLO_MobileNetV3_small | 4 | 32 | 16MB | 320 | 17.2 | 33.8 | 21.5 | [下载链接](https://paddledet.bj.bcebos.com/models/ppyolo_mbv3_small_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/ppyolo/ppyolo_mbv3_small_coco.yml) | | PP-YOLO_MobileNetV3_small | 4 | 32 | 16MB | 320 | 17.2 | 33.8 | 21.5 | [下载链接](https://paddledet.bj.bcebos.com/models/ppyolo_mbv3_small_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/configs/ppyolo/ppyolo_mbv3_small_coco.yml) |
- PP-YOLO_MobileNetV3 模型使用COCO数据集中train2017作为训练集，使用val2017作为测试集，Box AP<sup>val</sup>为`mAP(IoU=0.5:0.95)`评估结果, Box AP50<sup>val</sup>为`mAP(IoU=0.5)`评估结果。 - PP-YOLO_MobileNetV3 模型使用COCO数据集中train2017作为训练集，使用val2017作为测试集，Box AP<sup>val</sup>为`mAP(IoU=0.5:0.95)`评估结果, Box AP50<sup>val</sup>为`mAP(IoU=0.5)`评估结果。
- PP-YOLO_MobileNetV3 模型训练过程中使用4GPU，每GPU batch size为32进行训练，如训练GPU数和batch size不使用上述配置，须参考[FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/FAQ.md)调整学习率和迭代次数。 - PP-YOLO_MobileNetV3 模型训练过程中使用4GPU，每GPU batch size为32进行训练，如训练GPU数和batch size不使用上述配置，须参考[FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/docs/tutorials/FAQ.md)调整学习率和迭代次数。
- PP-YOLO_MobileNetV3 模型推理速度测试环境配置为麒麟990芯片单线程。 - PP-YOLO_MobileNetV3 模型推理速度测试环境配置为麒麟990芯片单线程。
### PP-YOLO tiny模型 ### PP-YOLO tiny模型
......
static/configs/acfpn/README.md
浏览文件 @ bf79506a
...@@ -13,4 +13,4 @@ Cao J, Chen Q, Guo J, et al. Attention-guided Context Feature Pyramid Network fo ...@@ -13,4 +13,4 @@ Cao J, Chen Q, Guo J, et al. Attention-guided Context Feature Pyramid Network fo
| Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-ACFPN | Faster | 2 | 1x | 23.432 | 39.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_acfpn_1x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/acfpn/faster_rcnn_r50_vd_acfpn_1x.yml) | | ResNet50-vd-ACFPN | Faster | 2 | 1x | 23.432 | 39.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_acfpn_1x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/acfpn/faster_rcnn_r50_vd_acfpn_1x.yml) |
static/configs/anchor_free/README.md
浏览文件 @ bf79506a
...@@ -26,14 +26,14 @@ ...@@ -26,14 +26,14 @@
| 网络结构 | 骨干网络 | 图片个数/GPU | 预训练模型 | mAP | FPS | 模型下载 | 配置文件 | | 网络结构 | 骨干网络 | 图片个数/GPU | 预训练模型 | mAP | FPS | 模型下载 | 配置文件 |
|:------------:|:--------:|:----:|:-------:|:-------:|:---------:|:----------:|:----------:| |:------------:|:--------:|:----:|:-------:|:-------:|:---------:|:----------:|:----------:|
| CornerNet-Squeeze | Hourglass104 | 14 | 无 | 34.5 | 35.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_hg104.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/cornernet_squeeze_hg104.yml) | | CornerNet-Squeeze | Hourglass104 | 14 | 无 | 34.5 | 35.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_hg104.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/cornernet_squeeze_hg104.yml) |
| CornerNet-Squeeze | ResNet50-vd | 14 | [faster\_rcnn\_r50\_vd\_fpn\_2x](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_2x.tar) | 32.7 | 47.01 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_r50_vd_fpn.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/cornernet_squeeze_r50_vd_fpn.yml) | | CornerNet-Squeeze | ResNet50-vd | 14 | [faster\_rcnn\_r50\_vd\_fpn\_2x](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_2x.tar) | 32.7 | 47.01 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_r50_vd_fpn.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/cornernet_squeeze_r50_vd_fpn.yml) |
| CornerNet-Squeeze-dcn | ResNet50-vd | 14 | [faster\_rcnn\_dcn\_r50\_vd\_fpn\_2x](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_2x.tar) | 34.9 | 40.43 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_dcn_r50_vd_fpn.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/cornernet_squeeze_dcn_r50_vd_fpn.yml) | | CornerNet-Squeeze-dcn | ResNet50-vd | 14 | [faster\_rcnn\_dcn\_r50\_vd\_fpn\_2x](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_2x.tar) | 34.9 | 40.43 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_dcn_r50_vd_fpn.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/cornernet_squeeze_dcn_r50_vd_fpn.yml) |
| CornerNet-Squeeze-dcn-mixup-cosine* | ResNet50-vd | 14 | [faster\_rcnn\_dcn\_r50\_vd\_fpn\_2x](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_2x.tar) | 38.2 | 39.70 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_dcn_r50_vd_fpn_mixup_cosine.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/cornernet_squeeze_dcn_r50_vd_fpn_mixup_cosine.yml) | | CornerNet-Squeeze-dcn-mixup-cosine* | ResNet50-vd | 14 | [faster\_rcnn\_dcn\_r50\_vd\_fpn\_2x](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_2x.tar) | 38.2 | 39.70 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cornernet_squeeze_dcn_r50_vd_fpn_mixup_cosine.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/cornernet_squeeze_dcn_r50_vd_fpn_mixup_cosine.yml) |
| FCOS | ResNet50 | 2 | [ResNet50\_cos\_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar) | 39.8 | 18.85 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/fcos_r50_fpn_1x.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/fcos_r50_fpn_1x.yml) | | FCOS | ResNet50 | 2 | [ResNet50\_cos\_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar) | 39.8 | 18.85 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/fcos_r50_fpn_1x.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/fcos_r50_fpn_1x.yml) |
| FCOS+multiscale_train | ResNet50 | 2 | [ResNet50\_cos\_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar) | 42.0 | 19.05 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/fcos_r50_fpn_multiscale_2x.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/fcos_r50_fpn_multiscale_2x.yml) | | FCOS+multiscale_train | ResNet50 | 2 | [ResNet50\_cos\_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar) | 42.0 | 19.05 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/fcos_r50_fpn_multiscale_2x.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/fcos_r50_fpn_multiscale_2x.yml) |
| FCOS+DCN | ResNet50 | 2 | [ResNet50\_cos\_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar) | 44.4 | 13.66 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/fcos_dcn_r50_fpn_1x.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/fcos_dcn_r50_fpn_1x.yml) | | FCOS+DCN | ResNet50 | 2 | [ResNet50\_cos\_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_cos_pretrained.tar) | 44.4 | 13.66 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/fcos_dcn_r50_fpn_1x.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/fcos_dcn_r50_fpn_1x.yml) |
| TTFNet | DarkNet53 | 12 | [DarkNet53_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/DarkNet53_pretrained.tar) | 32.9 | 85.92 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/ttfnet_darknet.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/anchor_free/ttfnet_darknet.yml) | | TTFNet | DarkNet53 | 12 | [DarkNet53_pretrained](https://paddle-imagenet-models-name.bj.bcebos.com/DarkNet53_pretrained.tar) | 32.9 | 85.92 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/ttfnet_darknet.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/anchor_free/ttfnet_darknet.yml) |
**注意:** **注意:**
......
static/configs/autoaugment/README.md
浏览文件 @ bf79506a
...@@ -19,5 +19,5 @@ ...@@ -19,5 +19,5 @@
| Backbone | Type | AutoAug policy | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | AutoAug policy | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------:| :-------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------:| :-------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-FPN | Faster | v1 | 2 | 3x | 22.800 | 39.9 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_aa_3x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/autoaugment/faster_rcnn_r50_vd_fpn_aa_3x.yml) | | ResNet50-vd-FPN | Faster | v1 | 2 | 3x | 22.800 | 39.9 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_aa_3x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/autoaugment/faster_rcnn_r50_vd_fpn_aa_3x.yml) |
| ResNet101-vd-FPN | Faster | v1 | 2 | 3x | 17.652 | 42.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r101_vd_fpn_aa_3x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/autoaugment/faster_rcnn_r101_vd_fpn_aa_3x.yml) | | ResNet101-vd-FPN | Faster | v1 | 2 | 3x | 17.652 | 42.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r101_vd_fpn_aa_3x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/autoaugment/faster_rcnn_r101_vd_fpn_aa_3x.yml) |
static/configs/gcnet/README.md
浏览文件 @ bf79506a
...@@ -30,5 +30,5 @@ ...@@ -30,5 +30,5 @@
| Backbone | Type | Context| Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Context| Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------: | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-FPN | Mask | GC(c3-c5, r16, add) | 2 | 2x | 15.31 | 41.4 | 36.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.yml) | | ResNet50-vd-FPN | Mask | GC(c3-c5, r16, add) | 2 | 2x | 15.31 | 41.4 | 36.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.yml) |
| ResNet50-vd-FPN | Mask | GC(c3-c5, r16, mul) | 2 | 2x | 15.35 | 40.7 | 36.1 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.yml) | | ResNet50-vd-FPN | Mask | GC(c3-c5, r16, mul) | 2 | 2x | 15.35 | 40.7 | 36.1 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.yml) |
static/configs/gcnet/README_cn.md
浏览文件 @ bf79506a
...@@ -42,8 +42,8 @@ SNL模块可以抽象为上下文建模、特征转换和特征聚合三个部 ...@@ -42,8 +42,8 @@ SNL模块可以抽象为上下文建模、特征转换和特征聚合三个部
| 骨架网络 | 网络类型 | Context设置 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 | | 骨架网络 | 网络类型 | Context设置 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 |
| :---------------------- | :-------------: | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-FPN | Mask | GC(c3-c5, r16, add) | 2 | 2x | 15.31 | 41.4 | 36.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.yml) | | ResNet50-vd-FPN | Mask | GC(c3-c5, r16, add) | 2 | 2x | 15.31 | 41.4 | 36.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_add_r16_2x.yml) |
| ResNet50-vd-FPN | Mask | GC(c3-c5, r16, mul) | 2 | 2x | 15.35 | 40.7 | 36.1 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.yml) | | ResNet50-vd-FPN | Mask | GC(c3-c5, r16, mul) | 2 | 2x | 15.35 | 40.7 | 36.1 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/gcnet/mask_rcnn_r50_vd_fpn_gcb_mul_r16_2x.yml) |
## 引用 ## 引用
......
static/configs/gridmask/README.md
浏览文件 @ bf79506a
...@@ -19,4 +19,4 @@ ...@@ -19,4 +19,4 @@
| Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-FPN | Faster | 2 | 4x | 21.847 | 39.1% | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_gridmask_4x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/gridmask/faster_rcnn_r50_vd_fpn_gridmask_4x.yml) | | ResNet50-vd-FPN | Faster | 2 | 4x | 21.847 | 39.1% | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_gridmask_4x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/gridmask/faster_rcnn_r50_vd_fpn_gridmask_4x.yml) |
static/configs/hrnet/README.md
浏览文件 @ bf79506a
...@@ -30,5 +30,5 @@ ...@@ -30,5 +30,5 @@
| Backbone | Type | deformable Conv | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | deformable Conv | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :------------- | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :------------- | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| HRNetV2p_W18 | Faster | False | 2 | 1x | 17.509 | 36.0 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_hrnetv2p_w18_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/hrnet/faster_rcnn_hrnetv2p_w18_1x.yml) | | HRNetV2p_W18 | Faster | False | 2 | 1x | 17.509 | 36.0 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_hrnetv2p_w18_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/hrnet/faster_rcnn_hrnetv2p_w18_1x.yml) |
| HRNetV2p_W18 | Faster | False | 2 | 2x | 17.509 | 38.0 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_hrnetv2p_w18_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/hrnet/faster_rcnn_hrnetv2p_w18_2x.yml) | | HRNetV2p_W18 | Faster | False | 2 | 2x | 17.509 | 38.0 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_hrnetv2p_w18_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/hrnet/faster_rcnn_hrnetv2p_w18_2x.yml) |
static/configs/iou_loss/README.md
浏览文件 @ bf79506a
...@@ -43,6 +43,6 @@ ...@@ -43,6 +43,6 @@
| Backbone | Type | Loss Type | Loss Weight | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Loss Type | Loss Weight | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :------------- | :---: | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :---: | | :---------------------- | :------------- | :---: | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :---: |
| ResNet50-vd-FPN | Faster | GIOU | 10 | 2 | 1x | 22.94 | 39.4 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_giou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/iou_loss/faster_rcnn_r50_vd_fpn_giou_loss_1x.yml) | | ResNet50-vd-FPN | Faster | GIOU | 10 | 2 | 1x | 22.94 | 39.4 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_giou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/iou_loss/faster_rcnn_r50_vd_fpn_giou_loss_1x.yml) |
| ResNet50-vd-FPN | Faster | DIOU | 12 | 2 | 1x | 22.94 | 39.2 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_diou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/iou_loss/faster_rcnn_r50_vd_fpn_diou_loss_1x.yml) | | ResNet50-vd-FPN | Faster | DIOU | 12 | 2 | 1x | 22.94 | 39.2 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_diou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/iou_loss/faster_rcnn_r50_vd_fpn_diou_loss_1x.yml) |
| ResNet50-vd-FPN | Faster | CIOU | 12 | 2 | 1x | 22.95 | 39.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_ciou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/iou_loss/faster_rcnn_r50_vd_fpn_ciou_loss_1x.yml) | | ResNet50-vd-FPN | Faster | CIOU | 12 | 2 | 1x | 22.95 | 39.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_ciou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/iou_loss/faster_rcnn_r50_vd_fpn_ciou_loss_1x.yml) |
static/configs/iou_loss/README_cn.md
浏览文件 @ bf79506a
...@@ -83,9 +83,9 @@ CIOU loss使得在边框回归时，与目标框有重叠甚至包含时能够 ...@@ -83,9 +83,9 @@ CIOU loss使得在边框回归时，与目标框有重叠甚至包含时能够
| 骨架网络 | 网络类型 | Loss类型 | Loss权重 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 | | 骨架网络 | 网络类型 | Loss类型 | Loss权重 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 |
| :---------------------- | :------------- | :---: | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :---: | | :---------------------- | :------------- | :---: | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :---: |
| ResNet50-vd-FPN | Faster | GIOU | 10 | 2 | 1x | 22.94 | 39.4 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_giou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/iou_loss/faster_rcnn_r50_vd_fpn_giou_loss_1x.yml) | | ResNet50-vd-FPN | Faster | GIOU | 10 | 2 | 1x | 22.94 | 39.4 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_giou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/iou_loss/faster_rcnn_r50_vd_fpn_giou_loss_1x.yml) |
| ResNet50-vd-FPN | Faster | DIOU | 12 | 2 | 1x | 22.94 | 39.2 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_diou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/iou_loss/faster_rcnn_r50_vd_fpn_diou_loss_1x.yml) | | ResNet50-vd-FPN | Faster | DIOU | 12 | 2 | 1x | 22.94 | 39.2 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_diou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/iou_loss/faster_rcnn_r50_vd_fpn_diou_loss_1x.yml) |
| ResNet50-vd-FPN | Faster | CIOU | 12 | 2 | 1x | 22.95 | 39.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_ciou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/iou_loss/faster_rcnn_r50_vd_fpn_ciou_loss_1x.yml) | | ResNet50-vd-FPN | Faster | CIOU | 12 | 2 | 1x | 22.95 | 39.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_ciou_loss_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/iou_loss/faster_rcnn_r50_vd_fpn_ciou_loss_1x.yml) |
......
static/configs/libra_rcnn/README.md
浏览文件 @ bf79506a
...@@ -19,5 +19,5 @@ ...@@ -19,5 +19,5 @@
| Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-BFP | Faster | 2 | 1x | 18.247 | 40.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r50_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/libra_rcnn/libra_rcnn_r50_vd_fpn_1x.yml) | | ResNet50-vd-BFP | Faster | 2 | 1x | 18.247 | 40.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r50_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/libra_rcnn/libra_rcnn_r50_vd_fpn_1x.yml) |
| ResNet101-vd-BFP | Faster | 2 | 1x | 14.865 | 42.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r101_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/libra_rcnn/libra_rcnn_r101_vd_fpn_1x.yml) | | ResNet101-vd-BFP | Faster | 2 | 1x | 14.865 | 42.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r101_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/libra_rcnn/libra_rcnn_r101_vd_fpn_1x.yml) |
static/configs/libra_rcnn/README_cn.md
浏览文件 @ bf79506a
...@@ -60,8 +60,8 @@ Faster RCNN中生成许多候选框之后，使用随机的方法挑选正负样 ...@@ -60,8 +60,8 @@ Faster RCNN中生成许多候选框之后，使用随机的方法挑选正负样
| 骨架网络 | 网络类型 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 | | 骨架网络 | 网络类型 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-BFP | Faster | 2 | 1x | 18.247 | 40.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r50_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/libra_rcnn/libra_rcnn_r50_vd_fpn_1x.yml) | | ResNet50-vd-BFP | Faster | 2 | 1x | 18.247 | 40.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r50_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/libra_rcnn/libra_rcnn_r50_vd_fpn_1x.yml) |
| ResNet101-vd-BFP | Faster | 2 | 1x | 14.865 | 42.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r101_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/libra_rcnn/libra_rcnn_r101_vd_fpn_1x.yml) | | ResNet101-vd-BFP | Faster | 2 | 1x | 14.865 | 42.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/libra_rcnn_r101_vd_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/libra_rcnn/libra_rcnn_r101_vd_fpn_1x.yml) |
## 引用 ## 引用
......
static/configs/ppyolo/README.md
浏览文件 @ bf79506a
...@@ -101,7 +101,7 @@ PP-YOLO improved performance and speed of YOLOv3 with following methods: ...@@ -101,7 +101,7 @@ PP-YOLO improved performance and speed of YOLOv3 with following methods:
**Notes:** **Notes:**
- PP-YOLO-tiny is trained on COCO train2017 datast and evaluated on val2017 dataset，Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5:0.95)`, Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5)`. - PP-YOLO-tiny is trained on COCO train2017 datast and evaluated on val2017 dataset，Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5:0.95)`, Box AP<sup>val</sup> is evaluation results of `mAP(IoU=0.5)`.
- PP-YOLO-tiny used 8 GPUs for training and mini-batch size as 32 on each GPU, if GPU number and mini-batch size is changed, learning rate and iteration times should be adjusted according [FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/FAQ.md). - PP-YOLO-tiny used 8 GPUs for training and mini-batch size as 32 on each GPU, if GPU number and mini-batch size is changed, learning rate and iteration times should be adjusted according [FAQ](https://github.com/PaddlePaddle/PaddleDetection/blob//develop/static/docs/FAQ.md).
- PP-YOLO-tiny inference speed is tested on Kirin 990 with 4 threads by arm8 - PP-YOLO-tiny inference speed is tested on Kirin 990 with 4 threads by arm8
- we alse provide PP-YOLO-tiny post quant inference model, which can compress model to **1.3MB** with nearly no inference on inference speed and performance - we alse provide PP-YOLO-tiny post quant inference model, which can compress model to **1.3MB** with nearly no inference on inference speed and performance
......
static/configs/random_erasing/README.md
浏览文件 @ bf79506a
...@@ -18,4 +18,4 @@ ...@@ -18,4 +18,4 @@
| Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-FPN | Faster | 2 | 4x | 21.847 | 39.0% | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_random_erasing_4x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/random_erasing/faster_rcnn_r50_vd_fpn_random_erasing_4x.yml) | | ResNet50-vd-FPN | Faster | 2 | 4x | 21.847 | 39.0% | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_r50_vd_fpn_random_erasing_4x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/random_erasing/faster_rcnn_r50_vd_fpn_random_erasing_4x.yml) |
static/configs/rcnn_enhance/README.md
浏览文件 @ bf79506a
...@@ -32,9 +32,9 @@ ...@@ -32,9 +32,9 @@
| 骨架网络 | 网络类型 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 | | 骨架网络 | 网络类型 | 每张GPU图片个数 | 学习率策略 |推理时间(fps) | Box AP | Mask AP | 下载 | 配置文件 |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :-------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :-------------: | :-----: |
| ResNet50-vd-FPN-Dcnv2 | Faster | 2 | 3x | 61.425 | 41.6 | - | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) | | ResNet50-vd-FPN-Dcnv2 | Faster | 2 | 3x | 61.425 | 41.6 | - | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) |
| ResNet50-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 20.001 | 47.8 | - | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) | | ResNet50-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 20.001 | 47.8 | - | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) |
| ResNet101-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 19.523 | 49.4 | - | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.yml) | | ResNet101-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 19.523 | 49.4 | - | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.yml) |
**注**：generic文件夹下面的配置文件对应的预训练模型均只支持预测，不支持训练与评估。 **注**：generic文件夹下面的配置文件对应的预训练模型均只支持预测，不支持训练与评估。
static/configs/rcnn_enhance/README_en.md
浏览文件 @ bf79506a
...@@ -36,9 +36,9 @@ And the following figure shows `mAP-Speed` curves for some common detectors. ...@@ -36,9 +36,9 @@ And the following figure shows `mAP-Speed` curves for some common detectors.
| Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :-------------: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| ResNet50-vd-FPN-Dcnv2 | Faster | 2 | 3x | 61.425 | 41.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) | | ResNet50-vd-FPN-Dcnv2 | Faster | 2 | 3x | 61.425 | 41.6 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/faster_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) |
| ResNet50-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 20.001 | 47.8 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) | | ResNet50-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 20.001 | 47.8 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/cascade_rcnn_dcn_r50_vd_fpn_3x_server_side.yml) |
| ResNet101-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 19.523 | 49.4 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.yml) | | ResNet101-vd-FPN-Dcnv2 | Cascade Faster | 2 | 3x | 19.523 | 49.4 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/cascade_rcnn_dcn_r101_vd_fpn_3x_server_side.yml) |
**Attention**: Pretrained models whose congigurations are in the directory `generic` just support inference but do not support training and evaluation as now. **Attention**: Pretrained models whose congigurations are in the directory `generic` just support inference but do not support training and evaluation as now.
static/configs/res2net/README.md
浏览文件 @ bf79506a
...@@ -30,7 +30,7 @@ ...@@ -30,7 +30,7 @@
| Backbone | Type | deformable Conv | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs | | Backbone | Type | deformable Conv | Image/gpu | Lr schd | Inf time (fps) | Box AP | Mask AP | Download | Configs |
| :---------------------- | :------------- | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: | | :---------------------- | :------------- | :---: | :-------: | :-----: | :------------: | :----: | :-----: | :----------------------------------------------------------: | :-----: |
| Res2Net50-FPN | Faster | False | 2 | 1x | 20.320 | 39.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_res2net50_vb_26w_4s_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/res2net/faster_rcnn_res2net50_vb_26w_4s_fpn_1x.yml) | | Res2Net50-FPN | Faster | False | 2 | 1x | 20.320 | 39.5 | - | [model](https://paddlemodels.bj.bcebos.com/object_detection/faster_rcnn_res2net50_vb_26w_4s_fpn_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/res2net/faster_rcnn_res2net50_vb_26w_4s_fpn_1x.yml) |
| Res2Net50-FPN | Mask | False | 2 | 2x | 16.069 | 40.7 | 36.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_res2net50_vb_26w_4s_fpn_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/res2net/faster_rcnn_res2net50_vb_26w_4s_fpn_2x.yml) | | Res2Net50-FPN | Mask | False | 2 | 2x | 16.069 | 40.7 | 36.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_res2net50_vb_26w_4s_fpn_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/res2net/faster_rcnn_res2net50_vb_26w_4s_fpn_2x.yml) |
| Res2Net50-vd-FPN | Mask | False | 2 | 2x | 15.816 | 40.9 | 36.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_res2net50_vd_26w_4s_fpn_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/res2net/mask_rcnn_res2net50_vd_26w_4s_fpn_2x.yml) | | Res2Net50-vd-FPN | Mask | False | 2 | 2x | 15.816 | 40.9 | 36.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_res2net50_vd_26w_4s_fpn_2x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/res2net/mask_rcnn_res2net50_vd_26w_4s_fpn_2x.yml) |
| Res2Net50-vd-FPN | Mask | True | 2 | 2x | 14.478 | 43.5 | 38.4 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_res2net50_vd_26w_4s_fpn_dcnv2_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/res2net/mask_rcnn_res2net50_vd_26w_4s_fpn_dcnv2_1x.yml) | | Res2Net50-vd-FPN | Mask | True | 2 | 2x | 14.478 | 43.5 | 38.4 | [model](https://paddlemodels.bj.bcebos.com/object_detection/mask_rcnn_res2net50_vd_26w_4s_fpn_dcnv2_1x.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/res2net/mask_rcnn_res2net50_vd_26w_4s_fpn_dcnv2_1x.yml) |
static/configs/solov2/README.md
浏览文件 @ bf79506a
...@@ -25,15 +25,15 @@ SOLOv2 (Segmenting Objects by Locations) is a fast instance segmentation framewo ...@@ -25,15 +25,15 @@ SOLOv2 (Segmenting Objects by Locations) is a fast instance segmentation framewo
| BlendMask | R50-FPN | True | 3x | 37.8 | 13.5 | V100 | - | - | | BlendMask | R50-FPN | True | 3x | 37.8 | 13.5 | V100 | - | - |
| SOLOv2 (Paper) | R50-FPN | False | 1x | 34.8 | 18.5 | V100 | - | - | | SOLOv2 (Paper) | R50-FPN | False | 1x | 34.8 | 18.5 | V100 | - | - |
| SOLOv2 (Paper) | X101-DCN-FPN | True | 3x | 42.4 | 5.9 | V100 | - | - | | SOLOv2 (Paper) | X101-DCN-FPN | True | 3x | 42.4 | 5.9 | V100 | - | - |
| SOLOv2 | Mobilenetv3-FPN | True | 3x | 30.0 | 50 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_mobilenetv3_fpn_448_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/solov2/solov2_mobilenetv3_fpn_448_3x.yml) | | SOLOv2 | Mobilenetv3-FPN | True | 3x | 30.0 | 50 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_mobilenetv3_fpn_448_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/solov2/solov2_mobilenetv3_fpn_448_3x.yml) |
| SOLOv2 | R50-FPN | False | 1x | 35.6 | 21.9 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_r50_fpn_1x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/solov2/solov2_r50_fpn_1x.yml) | | SOLOv2 | R50-FPN | False | 1x | 35.6 | 21.9 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_r50_fpn_1x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/solov2/solov2_r50_fpn_1x.yml) |
| SOLOv2 | R50-FPN | True | 3x | 37.9 | 21.9 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_r50_fpn_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/solov2/solov2_r50_fpn_3x.yml) | | SOLOv2 | R50-FPN | True | 3x | 37.9 | 21.9 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_r50_fpn_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/solov2/solov2_r50_fpn_3x.yml) |
| SOLOv2 | R101-VD-FPN | True | 3x | 42.6 | 12.1 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_r101_vd_fpn_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/solov2/solov2_r101_vd_fpn_3x.yml) | | SOLOv2 | R101-VD-FPN | True | 3x | 42.6 | 12.1 | V100 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_r101_vd_fpn_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/solov2/solov2_r101_vd_fpn_3x.yml) |
## Enhanced model ## Enhanced model
| Backbone | Input size | Lr schd | V100 FP32(FPS) | Mask AP<sup>val</sup> | Download | Configs | | Backbone | Input size | Lr schd | V100 FP32(FPS) | Mask AP<sup>val</sup> | Download | Configs |
| :---------------------: | :-------------------: | :-----: | :------------: | :-----: | :---------: | :------------------------: | | :---------------------: | :-------------------: | :-----: | :------------: | :-----: | :---------: | :------------------------: |
| Light-R50-VD-DCN-FPN | 512 | 3x | 38.6 | 38.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_light_r50_vd_fpn_dcn_512_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/solov2/solov2_light_r50_vd_fpn_dcn_512_3x.yml) | | Light-R50-VD-DCN-FPN | 512 | 3x | 38.6 | 38.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/solov2_light_r50_vd_fpn_dcn_512_3x.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/solov2/solov2_light_r50_vd_fpn_dcn_512_3x.yml) |
**Notes:** **Notes:**
......
static/configs/yolov4/README.md
浏览文件 @ bf79506a
...@@ -41,8 +41,8 @@ python tools/anchor_cluster.py -c ${config} -m ${method} -s ${size} ...@@ -41,8 +41,8 @@ python tools/anchor_cluster.py -c ${config} -m ${method} -s ${size}
| | GPU个数 | 测试集 | 骨干网络 | 精度 | 模型下载 | 配置文件 | | | GPU个数 | 测试集 | 骨干网络 | 精度 | 模型下载 | 配置文件 |
|:------------------------:|:-------:|:------:|:--------------------------:|:------------------------:| :---------:| :-----: | |:------------------------:|:-------:|:------:|:--------------------------:|:------------------------:| :---------:| :-----: |
| YOLO v4 | - |test-dev2019 | CSPDarkNet53 | 43.5 |[下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov4_cspdarknet.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov4/yolov4_cspdarknet.yml) | | YOLO v4 | - |test-dev2019 | CSPDarkNet53 | 43.5 |[下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov4_cspdarknet.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/yolov4/yolov4_cspdarknet.yml) |
| YOLO v4 VOC | 2 | VOC2007 | CSPDarkNet53 | 85.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov4_cspdarknet_voc.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov4/yolov4_cspdarknet_voc.yml) | | YOLO v4 VOC | 2 | VOC2007 | CSPDarkNet53 | 85.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov4_cspdarknet_voc.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/yolov4/yolov4_cspdarknet_voc.yml) |
**注意:** **注意:**
......
static/deploy/README.md
浏览文件 @ bf79506a
...@@ -8,7 +8,7 @@ ...@@ -8,7 +8,7 @@
## 模型导出 ## 模型导出
训练得到一个满足要求的模型后，如果想要将该模型接入到C++服务器端预测库或移动端预测库，需要通过`tools/export_model.py`导出该模型。 训练得到一个满足要求的模型后，如果想要将该模型接入到C++服务器端预测库或移动端预测库，需要通过`tools/export_model.py`导出该模型。
- [导出教程](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/advanced_tutorials/deploy/EXPORT_MODEL.md) - [导出教程](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)
模型导出后, 目录结构如下(以`yolov3_darknet`为例): 模型导出后, 目录结构如下(以`yolov3_darknet`为例):
``` ```
...@@ -21,8 +21,8 @@ yolov3_darknet # 模型目录 ...@@ -21,8 +21,8 @@ yolov3_darknet # 模型目录
预测时，该目录所在的路径会作为程序的输入参数。 预测时，该目录所在的路径会作为程序的输入参数。
## 预测部署 ## 预测部署
- [1. Python预测(支持 Linux 和 Windows)](https://github.com/PaddlePaddle/PaddleDetection/blob/master/deploy/python) - [1. Python预测(支持 Linux 和 Windows)](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/deploy/python)
- [2. C++预测(支持 Linux 和 Windows)](https://github.com/PaddlePaddle/PaddleDetection/blob/master/deploy/cpp) - [2. C++预测(支持 Linux 和 Windows)](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/deploy/cpp)
- [3. 在线服务化部署](./serving/README.md) - [3. 在线服务化部署](./serving/README.md)
- [4. 移动端部署](https://github.com/PaddlePaddle/Paddle-Lite-Demo) - [4. 移动端部署](https://github.com/PaddlePaddle/Paddle-Lite-Demo)
- [5. Jetson设备部署](./cpp/docs/Jetson_build.md) - [5. Jetson设备部署](./cpp/docs/Jetson_build.md)
static/deploy/cpp/README.md
浏览文件 @ bf79506a
...@@ -52,7 +52,7 @@ deploy/cpp ...@@ -52,7 +52,7 @@ deploy/cpp
## 3.编译部署 ## 3.编译部署
### 3.1 导出模型 ### 3.1 导出模型
请确认您已经基于`PaddleDetection`的[export_model.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/export_model.py)导出您的模型，并妥善保存到合适的位置。导出模型细节请参考 [导出模型教程](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)。 请确认您已经基于`PaddleDetection`的[export_model.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/export_model.py)导出您的模型，并妥善保存到合适的位置。导出模型细节请参考 [导出模型教程](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)。
模型导出后, 目录结构如下(以`yolov3_darknet`为例): 模型导出后, 目录结构如下(以`yolov3_darknet`为例):
``` ```
......
static/deploy/python/README.md
浏览文件 @ bf79506a
...@@ -3,7 +3,7 @@ ...@@ -3,7 +3,7 @@
Python预测可以使用`tools/infer.py`，此种方式依赖PaddleDetection源码；也可以使用本篇教程预测方式，先将模型导出，使用一个独立的文件进行预测。 Python预测可以使用`tools/infer.py`，此种方式依赖PaddleDetection源码；也可以使用本篇教程预测方式，先将模型导出，使用一个独立的文件进行预测。
本篇教程使用AnalysisPredictor对[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)进行高性能预测。 本篇教程使用AnalysisPredictor对[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)进行高性能预测。
在PaddlePaddle中预测引擎和训练引擎底层有着不同的优化方法, 下面列出了两种不同的预测方式。Executor同时支持训练和预测，AnalysisPredictor则专门针对推理进行了优化，是基于[C++预测库](https://www.paddlepaddle.org.cn/documentation/docs/zh/advanced_guide/inference_deployment/inference/native_infer.html)的Python接口，该引擎可以对模型进行多项图优化，减少不必要的内存拷贝。如果用户在部署已训练模型的过程中对性能有较高的要求，我们提供了独立于PaddleDetection的预测脚本，方便用户直接集成部署。 在PaddlePaddle中预测引擎和训练引擎底层有着不同的优化方法, 下面列出了两种不同的预测方式。Executor同时支持训练和预测，AnalysisPredictor则专门针对推理进行了优化，是基于[C++预测库](https://www.paddlepaddle.org.cn/documentation/docs/zh/advanced_guide/inference_deployment/inference/native_infer.html)的Python接口，该引擎可以对模型进行多项图优化，减少不必要的内存拷贝。如果用户在部署已训练模型的过程中对性能有较高的要求，我们提供了独立于PaddleDetection的预测脚本，方便用户直接集成部署。
...@@ -18,7 +18,7 @@ Python预测可以使用`tools/infer.py`，此种方式依赖PaddleDetection源 ...@@ -18,7 +18,7 @@ Python预测可以使用`tools/infer.py`，此种方式依赖PaddleDetection源
## 1. 导出预测模型 ## 1. 导出预测模型
PaddleDetection在训练过程包括网络的前向和优化器相关参数，而在部署过程中，我们只需要前向参数，具体参考:[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/advanced_tutorials/deploy/EXPORT_MODEL.md) PaddleDetection在训练过程包括网络的前向和优化器相关参数，而在部署过程中，我们只需要前向参数，具体参考:[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)
导出后目录下，包括`__model__`，`__params__`和`infer_cfg.yml`三个文件。 导出后目录下，包括`__model__`，`__params__`和`infer_cfg.yml`三个文件。
......
static/deploy/serving/README.md
浏览文件 @ bf79506a
...@@ -22,7 +22,7 @@ pip install paddle-serving-server-gpu -i https://mirror.baidu.com/pypi/simple ...@@ -22,7 +22,7 @@ pip install paddle-serving-server-gpu -i https://mirror.baidu.com/pypi/simple
``` ```
## 3. 导出模型 ## 3. 导出模型
PaddleDetection在训练过程包括网络的前向和优化器相关参数，而在部署过程中，我们只需要前向参数，具体参考:[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/advanced_tutorials/deploy/EXPORT_MODEL.md) PaddleDetection在训练过程包括网络的前向和优化器相关参数，而在部署过程中，我们只需要前向参数，具体参考:[导出模型](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/advanced_tutorials/deploy/EXPORT_MODEL.md)
``` ```
python tools/export_serving_model.py -c configs/yolov3_mobilenet_v1_roadsign.yml -o use_gpu=true weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_roadsign.pdparams --output_dir=./inference_model python tools/export_serving_model.py -c configs/yolov3_mobilenet_v1_roadsign.yml -o use_gpu=true weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_roadsign.pdparams --output_dir=./inference_model
......
static/docs/MODEL_ZOO.md
浏览文件 @ bf79506a
此差异已折叠。
static/docs/MODEL_ZOO_cn.md
浏览文件 @ bf79506a
此差异已折叠。
static/docs/advanced_tutorials/MODEL_TECHNICAL.md
浏览文件 @ bf79506a
...@@ -34,7 +34,7 @@ PaddleDetection的网络模型模块所有代码逻辑在`ppdet/modeling/`中， ...@@ -34,7 +34,7 @@ PaddleDetection的网络模型模块所有代码逻辑在`ppdet/modeling/`中，
![](../images/models_figure.png) ![](../images/models_figure.png)
## 新增模型 ## 新增模型
我们以单阶段检测器YOLOv3为例，结合[yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml)配置文件，对建立模型过程进行详细描述， 我们以单阶段检测器YOLOv3为例，结合[yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml)配置文件，对建立模型过程进行详细描述，
按照此思路您可以快速搭建新的模型。 按照此思路您可以快速搭建新的模型。
搭建新模型的一般步骤是：Backbone编写、检测组件编写与模型组网这三个步骤，下面为您详细介绍： 搭建新模型的一般步骤是：Backbone编写、检测组件编写与模型组网这三个步骤，下面为您详细介绍：
......
static/docs/advanced_tutorials/READER.md
浏览文件 @ bf79506a
...@@ -392,7 +392,7 @@ EvalReader： ...@@ -392,7 +392,7 @@ EvalReader：
`DataLoader`的API详见[fluid.io.DataLoader](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api_cn/io_cn/DataLoader_cn.html#dataloader)。 `DataLoader`的API详见[fluid.io.DataLoader](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/api_cn/io_cn/DataLoader_cn.html#dataloader)。
具体步骤如下： 具体步骤如下：
- 在[train.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/train.py)、[eval.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/eval.py)和[infer.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/infer.py)里创建训练时的Reader： - 在[train.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/train.py)、[eval.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/eval.py)和[infer.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/infer.py)里创建训练时的Reader：
```python ```python
# 创建DataLoader对象 # 创建DataLoader对象
inputs_def = cfg['TestReader']['inputs_def'] inputs_def = cfg['TestReader']['inputs_def']
......
static/docs/advanced_tutorials/TRANSFER_LEARNING.md
浏览文件 @ bf79506a
...@@ -11,7 +11,7 @@ In transfer learning, if different dataset and the number of classes is used, th ...@@ -11,7 +11,7 @@ In transfer learning, if different dataset and the number of classes is used, th
Transfer learning needs custom dataset and annotation in COCO-format and VOC-format is supported now. The script converts the annotation from voc, labelme or cityscape to COCO is provided in ```tools/x2coco.py```. More details please refer to [READER](READER.md). After data preparation, update the data parameters in configuration file. Transfer learning needs custom dataset and annotation in COCO-format and VOC-format is supported now. The script converts the annotation from voc, labelme or cityscape to COCO is provided in ```tools/x2coco.py```. More details please refer to [READER](READER.md). After data preparation, update the data parameters in configuration file.
1. COCO-format dataset, take [yolov3\_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml#L66) for example, modify the COCODataSet in yolov3\_reader: 1. COCO-format dataset, take [yolov3\_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml#L66) for example, modify the COCODataSet in yolov3\_reader:
```yml ```yml
dataset: dataset:
...@@ -22,7 +22,7 @@ Transfer learning needs custom dataset and annotation in COCO-format and VOC-for ...@@ -22,7 +22,7 @@ Transfer learning needs custom dataset and annotation in COCO-format and VOC-for
with_background: false with_background: false
``` ```
2. VOC-format dataset, take [yolov3\_darknet\_voc.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet_voc.yml#L67) for example, modify the VOCDataSet in the configuration: 2. VOC-format dataset, take [yolov3\_darknet\_voc.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet_voc.yml#L67) for example, modify the VOCDataSet in the configuration:
```yml ```yml
dataset: dataset:
...@@ -52,7 +52,7 @@ python -u tools/train.py -c configs/faster_rcnn_r50_1x.yml \ ...@@ -52,7 +52,7 @@ python -u tools/train.py -c configs/faster_rcnn_r50_1x.yml \
The parameters which need to ignore can be specified explicitly as well and arbitrary parameter names can be added to `finetune_exclude_pretrained_params`. For this purpose, several methods can be used as follwed: The parameters which need to ignore can be specified explicitly as well and arbitrary parameter names can be added to `finetune_exclude_pretrained_params`. For this purpose, several methods can be used as follwed:
- Set `finetune_exclude_pretrained_params` in YAML configuration files. Please refer to [configure file](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_mobilenet_v1_fruit.yml#L15) - Set `finetune_exclude_pretrained_params` in YAML configuration files. Please refer to [configure file](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_mobilenet_v1_fruit.yml#L15)
- Set `finetune_exclude_pretrained_params` in command line. For example: - Set `finetune_exclude_pretrained_params` in command line. For example:
```python ```python
......
static/docs/advanced_tutorials/TRANSFER_LEARNING_cn.md
浏览文件 @ bf79506a
...@@ -9,7 +9,7 @@ ...@@ -9,7 +9,7 @@
迁移学习需要使用自己的数据集，目前已支持COCO和VOC的数据标注格式，在```tools/x2coco.py```中给出了voc、labelme和cityscape标注格式转换为COCO格式的脚本，具体使用方式可以参考[自定义数据源](READER.md)。数据准备完成后，在配置文件中配置数据路径，对应修改reader中的路径参数即可。 迁移学习需要使用自己的数据集，目前已支持COCO和VOC的数据标注格式，在```tools/x2coco.py```中给出了voc、labelme和cityscape标注格式转换为COCO格式的脚本，具体使用方式可以参考[自定义数据源](READER.md)。数据准备完成后，在配置文件中配置数据路径，对应修改reader中的路径参数即可。
1. COCO数据集需要修改COCODataSet中的参数，以[yolov3\_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml#L66)为例，修改yolov3\_reader中的配置： 1. COCO数据集需要修改COCODataSet中的参数，以[yolov3\_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml#L66)为例，修改yolov3\_reader中的配置：
```yml ```yml
dataset: dataset:
...@@ -20,7 +20,7 @@ ...@@ -20,7 +20,7 @@
with_background: false with_background: false
``` ```
2. VOC数据集需要修改VOCDataSet中的参数，以[yolov3\_darknet\_voc.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet_voc.yml#L67)为例： 2. VOC数据集需要修改VOCDataSet中的参数，以[yolov3\_darknet\_voc.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet_voc.yml#L67)为例：
```yml ```yml
dataset: dataset:
...@@ -56,7 +56,7 @@ python -u tools/train.py -c configs/faster_rcnn_r50_1x.yml \ ...@@ -56,7 +56,7 @@ python -u tools/train.py -c configs/faster_rcnn_r50_1x.yml \
可以显示的指定训练过程中忽略参数的名字，任何参数名均可加入`finetune_exclude_pretrained_params`中，为实现这一目的，可通过如下方式实现： 可以显示的指定训练过程中忽略参数的名字，任何参数名均可加入`finetune_exclude_pretrained_params`中，为实现这一目的，可通过如下方式实现：
1. 在 YMAL 配置文件中通过设置`finetune_exclude_pretrained_params`字段。可参考[配置文件](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_mobilenet_v1_fruit.yml#L15) 1. 在 YMAL 配置文件中通过设置`finetune_exclude_pretrained_params`字段。可参考[配置文件](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_mobilenet_v1_fruit.yml#L15)
2. 在 train.py的启动参数中设置`finetune_exclude_pretrained_params`。例如： 2. 在 train.py的启动参数中设置`finetune_exclude_pretrained_params`。例如：
```python ```python
......
static/docs/advanced_tutorials/deploy/docs/windows_vs2015_build.md
浏览文件 @ bf79506a
# Windows平台使用 Visual Studio 2015 编译指南 # Windows平台使用 Visual Studio 2015 编译指南
本文档步骤，我们同时在`Visual Studio 2015` 和 `Visual Studio 2019 Community` 两个版本进行了测试，我们推荐使用[`Visual Studio 2019`直接编译`CMake`项目](https://github.com/PaddlePaddle/PaddleDetection/blob/master/deploy/cpp/docs/windows_vs2019_build.md)。 本文档步骤，我们同时在`Visual Studio 2015` 和 `Visual Studio 2019 Community` 两个版本进行了测试，我们推荐使用[`Visual Studio 2019`直接编译`CMake`项目](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/deploy/cpp/docs/windows_vs2019_build.md)。
## 前置条件 ## 前置条件
......
static/docs/conf.py
浏览文件 @ bf79506a
...@@ -162,7 +162,7 @@ texinfo_documents = [ ...@@ -162,7 +162,7 @@ texinfo_documents = [
def url_resolver(url): def url_resolver(url):
if ".html" not in url: if ".html" not in url:
url = url.replace("../", "") url = url.replace("../", "")
return "https://github.com/PaddlePaddle/PaddleDetection/tree/master" + url return "https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static" + url
else: else:
if DEPLOY: if DEPLOY:
return "http://paddledetection.readthedocs.io/" + url return "http://paddledetection.readthedocs.io/" + url
......
static/docs/featured_model/CONTRIB.md
浏览文件 @ bf79506a
...@@ -5,8 +5,8 @@ We provide some models implemented by PaddlePaddle to detect objects in specific ...@@ -5,8 +5,8 @@ We provide some models implemented by PaddlePaddle to detect objects in specific
| Task | Algorithm | Box AP | Download | Configs | | Task | Algorithm | Box AP | Download | Configs |
|:---------------------|:---------:|:------:| :-------------------------------------------------------------------------------------: |:------:| |:---------------------|:---------:|:------:| :-------------------------------------------------------------------------------------: |:------:|
| Vehicle Detection | YOLOv3 | 54.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/vehicle_yolov3_darknet.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/contrib/VehicleDetection/vehicle_yolov3_darknet.yml) | | Vehicle Detection | YOLOv3 | 54.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/vehicle_yolov3_darknet.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/contrib/VehicleDetection/vehicle_yolov3_darknet.yml) |
| Pedestrian Detection | YOLOv3 | 51.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/pedestrian_yolov3_darknet.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/contrib/PedestrianDetection/pedestrian_yolov3_darknet.yml) | | Pedestrian Detection | YOLOv3 | 51.8 | [model](https://paddlemodels.bj.bcebos.com/object_detection/pedestrian_yolov3_darknet.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/contrib/PedestrianDetection/pedestrian_yolov3_darknet.yml) |
## Vehicle Detection ## Vehicle Detection
...@@ -18,7 +18,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53 ...@@ -18,7 +18,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53
### 2. Configuration for training ### 2. Configuration for training
PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for vehicle detection: PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for vehicle detection:
* max_iters: 120000 * max_iters: 120000
* num_classes: 6 * num_classes: 6
...@@ -67,7 +67,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53 ...@@ -67,7 +67,7 @@ The network for detecting vehicles is YOLOv3, the backbone of which is Dacknet53
### 2. Configuration for training ### 2. Configuration for training
PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for pedestrian detection: PaddleDetection provides users with a configuration file [yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml) to train YOLOv3 on the COCO dataset, compared with this file, we modify some parameters as followed to conduct the training for pedestrian detection:
* max_iters: 200000 * max_iters: 200000
* num_classes: 1 * num_classes: 1
......
static/docs/featured_model/CONTRIB_cn.md
浏览文件 @ bf79506a
...@@ -5,8 +5,8 @@ ...@@ -5,8 +5,8 @@
| 任务 | 算法 | 精度(Box AP) | 下载 | 配置文件 | | 任务 | 算法 | 精度(Box AP) | 下载 | 配置文件 |
|:---------------------|:---------:|:------:| :---------------------------------------------------------------------------------: | :------:| |:---------------------|:---------:|:------:| :---------------------------------------------------------------------------------: | :------:|
| 车辆检测 | YOLOv3 | 54.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/vehicle_yolov3_darknet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/contrib/VehicleDetection/vehicle_yolov3_darknet.yml) | | 车辆检测 | YOLOv3 | 54.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/vehicle_yolov3_darknet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/contrib/VehicleDetection/vehicle_yolov3_darknet.yml) |
| 行人检测 | YOLOv3 | 51.8 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/pedestrian_yolov3_darknet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/contrib/PedestrianDetection/pedestrian_yolov3_darknet.yml) | | 行人检测 | YOLOv3 | 51.8 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/pedestrian_yolov3_darknet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/contrib/PedestrianDetection/pedestrian_yolov3_darknet.yml) |
## 车辆检测（Vehicle Detection） ## 车辆检测（Vehicle Detection）
...@@ -19,7 +19,7 @@ Backbone为Dacknet53的YOLOv3。 ...@@ -19,7 +19,7 @@ Backbone为Dacknet53的YOLOv3。
### 2. 训练参数配置 ### 2. 训练参数配置
PaddleDetection提供了使用COCO数据集对YOLOv3进行训练的参数配置文件[yolov3_darnet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml)，与之相比，在进行车辆检测的模型训练时，我们对以下参数进行了修改： PaddleDetection提供了使用COCO数据集对YOLOv3进行训练的参数配置文件[yolov3_darnet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml)，与之相比，在进行车辆检测的模型训练时，我们对以下参数进行了修改：
* max_iters: 120000 * max_iters: 120000
* num_classes: 6 * num_classes: 6
...@@ -69,7 +69,7 @@ Backbone为Dacknet53的YOLOv3。 ...@@ -69,7 +69,7 @@ Backbone为Dacknet53的YOLOv3。
### 2. 训练参数配置 ### 2. 训练参数配置
PaddleDetection提供了使用COCO数据集对YOLOv3进行训练的参数配置文件[yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/yolov3_darknet.yml)，与之相比，在进行行人检测的模型训练时，我们对以下参数进行了修改： PaddleDetection提供了使用COCO数据集对YOLOv3进行训练的参数配置文件[yolov3_darknet.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/yolov3_darknet.yml)，与之相比，在进行行人检测的模型训练时，我们对以下参数进行了修改：
* max_iters: 200000 * max_iters: 200000
* num_classes: 1 * num_classes: 1
......
static/docs/featured_model/FACE_DETECTION.md
浏览文件 @ bf79506a
...@@ -34,16 +34,16 @@ FaceDetection的目标是提供高效、高速的人脸检测解决方案，包 ...@@ -34,16 +34,16 @@ FaceDetection的目标是提供高效、高速的人脸检测解决方案，包
| 网络结构 | 类型 | 输入尺寸 | 图片个数/GPU | 学习率策略 | Easy Set | Medium Set | Hard Set | 下载 | 配置文件 | | 网络结构 | 类型 | 输入尺寸 | 图片个数/GPU | 学习率策略 | Easy Set | Medium Set | Hard Set | 下载 | 配置文件 |
|:------------:|:--------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:| |:------------:|:--------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:|
| BlazeFace | 原始版本 | 640 | 8 | 32w | **0.915** | **0.892** | **0.797** | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_original.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface.yml) | | BlazeFace | 原始版本 | 640 | 8 | 32w | **0.915** | **0.892** | **0.797** | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_original.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface.yml) |
| BlazeFace | Lite版本 | 640 | 8 | 32w | 0.909 | 0.885 | 0.781 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_lite.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface.yml) | | BlazeFace | Lite版本 | 640 | 8 | 32w | 0.909 | 0.885 | 0.781 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_lite.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface.yml) |
| BlazeFace | NAS版本 | 640 | 8 | 32w | 0.837 | 0.807 | 0.658 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface_nas.yml) | | BlazeFace | NAS版本 | 640 | 8 | 32w | 0.837 | 0.807 | 0.658 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface_nas.yml) |
| BlazeFace | NAS_V2版本 | 640 | 8 | 32W | 0.870 | 0.837 | 0.685 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas2.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface_nas_v2.yml) | | BlazeFace | NAS_V2版本 | 640 | 8 | 32W | 0.870 | 0.837 | 0.685 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas2.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface_nas_v2.yml) |
| FaceBoxes | 原始版本 | 640 | 8 | 32w | 0.878 | 0.851 | 0.576 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_original.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/faceboxes.yml) | | FaceBoxes | 原始版本 | 640 | 8 | 32w | 0.878 | 0.851 | 0.576 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_original.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/faceboxes.yml) |
| FaceBoxes | Lite版本 | 640 | 8 | 32w | 0.901 | 0.875 | 0.760 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_lite.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/faceboxes_lite.yml) | | FaceBoxes | Lite版本 | 640 | 8 | 32w | 0.901 | 0.875 | 0.760 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_lite.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/faceboxes_lite.yml) |
**注意:** **注意:**
- 我们使用`tools/face_eval.py`中多尺度评估策略得到`Easy/Medium/Hard Set`里的mAP。具体细节请参考[在WIDER-FACE数据集上评估](#在WIDER-FACE数据集上评估)。 - 我们使用`tools/face_eval.py`中多尺度评估策略得到`Easy/Medium/Hard Set`里的mAP。具体细节请参考[在WIDER-FACE数据集上评估](#在WIDER-FACE数据集上评估)。
- BlazeFace-Lite的训练与测试使用 [blazeface.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/face_detection/blazeface.yml)配置文件并且设置：`lite_edition: true`。 - BlazeFace-Lite的训练与测试使用 [blazeface.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/face_detection/blazeface.yml)配置文件并且设置：`lite_edition: true`。
#### FDDB数据集上的mAP #### FDDB数据集上的mAP
...@@ -258,7 +258,7 @@ wget https://dataset.bj.bcebos.com/wider_face/wider_face_train_bbx_lmk_gt.txt ...@@ -258,7 +258,7 @@ wget https://dataset.bj.bcebos.com/wider_face/wider_face_train_bbx_lmk_gt.txt
| 网络结构 | 输入尺寸 | 图片个数/GPU | 学习率策略 | Easy Set | Medium Set | Hard Set | 下载 | 配置文件 | | 网络结构 | 输入尺寸 | 图片个数/GPU | 学习率策略 | Easy Set | Medium Set | Hard Set | 下载 | 配置文件 |
|:------------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:| |:------------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:|
| BlazeFace Keypoint | 640 | 16 | 16w | 0.852 | 0.816 | 0.662 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_keypoint.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface_keypoint.yml) | | BlazeFace Keypoint | 640 | 16 | 16w | 0.852 | 0.816 | 0.662 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_keypoint.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface_keypoint.yml) |
![](../images/12_Group_Group_12_Group_Group_12_84.jpg) ![](../images/12_Group_Group_12_Group_Group_12_84.jpg)
...@@ -292,7 +292,7 @@ wget https://dataset.bj.bcebos.com/wider_face/wider_face_train_bbx_lmk_gt.txt ...@@ -292,7 +292,7 @@ wget https://dataset.bj.bcebos.com/wider_face/wider_face_train_bbx_lmk_gt.txt
**版本信息:** **版本信息:**
- 原始版本: 参考原始论文进行修改； - 原始版本: 参考原始论文进行修改；
- Lite版本: 使用更少的网络层数和通道数，具体可参考[代码](https://github.com/PaddlePaddle/PaddleDetection/blob/master/ppdet/modeling/architectures/faceboxes.py)。 - Lite版本: 使用更少的网络层数和通道数，具体可参考[代码](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/ppdet/modeling/architectures/faceboxes.py)。
## 如何贡献代码 ## 如何贡献代码
......
static/docs/featured_model/FACE_DETECTION_en.md
浏览文件 @ bf79506a
...@@ -38,17 +38,17 @@ optimized network structure. ...@@ -38,17 +38,17 @@ optimized network structure.
| Architecture | Type | Size | Img/gpu | Lr schd | Easy Set | Medium Set | Hard Set | Download | Configs | | Architecture | Type | Size | Img/gpu | Lr schd | Easy Set | Medium Set | Hard Set | Download | Configs |
|:------------:|:--------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:| |:------------:|:--------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:|
| BlazeFace | Original | 640 | 8 | 32w | **0.915** | **0.892** | **0.797** | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_original.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface.yml) | | BlazeFace | Original | 640 | 8 | 32w | **0.915** | **0.892** | **0.797** | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_original.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface.yml) |
| BlazeFace | Lite | 640 | 8 | 32w | 0.909 | 0.885 | 0.781 | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_lite.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface.yml) | | BlazeFace | Lite | 640 | 8 | 32w | 0.909 | 0.885 | 0.781 | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_lite.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface.yml) |
| BlazeFace | NAS | 640 | 8 | 32w | 0.837 | 0.807 | 0.658 | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface_nas.yml) | | BlazeFace | NAS | 640 | 8 | 32w | 0.837 | 0.807 | 0.658 | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface_nas.yml) |
| BlazeFace | NAS_V2 | 640 | 8 | 32W | 0.870 | 0.837 | 0.685 | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas2.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface_nas_v2.yml) | | BlazeFace | NAS_V2 | 640 | 8 | 32W | 0.870 | 0.837 | 0.685 | [model](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_nas2.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface_nas_v2.yml) |
| FaceBoxes | Original | 640 | 8 | 32w | 0.878 | 0.851 | 0.576 | [model](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_original.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/faceboxes.yml) | | FaceBoxes | Original | 640 | 8 | 32w | 0.878 | 0.851 | 0.576 | [model](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_original.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/faceboxes.yml) |
| FaceBoxes | Lite | 640 | 8 | 32w | 0.901 | 0.875 | 0.760 | [model](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_lite.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/faceboxes_lite.yml) | | FaceBoxes | Lite | 640 | 8 | 32w | 0.901 | 0.875 | 0.760 | [model](https://paddlemodels.bj.bcebos.com/object_detection/faceboxes_lite.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/faceboxes_lite.yml) |
**NOTES:** **NOTES:**
- Get mAP in `Easy/Medium/Hard Set` by multi-scale evaluation in `tools/face_eval.py`. - Get mAP in `Easy/Medium/Hard Set` by multi-scale evaluation in `tools/face_eval.py`.
For details can refer to [Evaluation](#Evaluate-on-the-WIDER-FACE). For details can refer to [Evaluation](#Evaluate-on-the-WIDER-FACE).
- BlazeFace-Lite Training and Testing ues [blazeface.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/face_detection/blazeface.yml) - BlazeFace-Lite Training and Testing ues [blazeface.yml](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/face_detection/blazeface.yml)
configs file and set `lite_edition: true`. configs file and set `lite_edition: true`.
#### mAP in FDDB #### mAP in FDDB
...@@ -274,7 +274,7 @@ wget https://dataset.bj.bcebos.com/wider_face/wider_face_train_bbx_lmk_gt.txt ...@@ -274,7 +274,7 @@ wget https://dataset.bj.bcebos.com/wider_face/wider_face_train_bbx_lmk_gt.txt
| Architecture | Size | Img/gpu | Lr schd | Easy Set | Medium Set | Hard Set | Download | Configs | | Architecture | Size | Img/gpu | Lr schd | Easy Set | Medium Set | Hard Set | Download | Configs |
|:------------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:| |:------------:|:----:|:-------:|:-------:|:---------:|:----------:|:---------:|:--------:|:--------:|
| BlazeFace Keypoint | 640 | 16 | 16w | 0.852 | 0.816 | 0.662 | [download](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_keypoint.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/face_detection/blazeface_keypoint.yml) | | BlazeFace Keypoint | 640 | 16 | 16w | 0.852 | 0.816 | 0.662 | [download](https://paddlemodels.bj.bcebos.com/object_detection/blazeface_keypoint.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/face_detection/blazeface_keypoint.yml) |
![](../images/12_Group_Group_12_Group_Group_12_84.jpg) ![](../images/12_Group_Group_12_Group_Group_12_84.jpg)
......
static/docs/featured_model/LARGE_SCALE_DET_MODEL.md
浏览文件 @ bf79506a
...@@ -20,6 +20,6 @@ ...@@ -20,6 +20,6 @@
| 骨架网络 | 网络类型 | 下载 | 配置文件 | | 骨架网络 | 网络类型 | 下载 | 配置文件 |
| :---------------| :---------------| :---------------| :--------------- | :---------------| :---------------| :---------------| :---------------
| ResNet50-vd-FPN-Dcnv2 | Cascade Faster | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.yml) | | ResNet50-vd-FPN-Dcnv2 | Cascade Faster | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.yml) |
| ResNet101-vd-FPN-Dcnv2 | Cascade Faster | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.yml) | | ResNet101-vd-FPN-Dcnv2 | Cascade Faster | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.yml) |
| CBResNet101-vd-FPN | Cascade Faster | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_cbr101_vd_fpn_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/generic/cascade_rcnn_cbr101_vd_fpn_server_side.yml) | | CBResNet101-vd-FPN | Cascade Faster | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_cbr101_vd_fpn_server_side.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/generic/cascade_rcnn_cbr101_vd_fpn_server_side.yml) |
static/docs/featured_model/LARGE_SCALE_DET_MODEL_en.md
浏览文件 @ bf79506a
...@@ -18,6 +18,6 @@ There are relatively not enough categories in the above dataset (compared to 100 ...@@ -18,6 +18,6 @@ There are relatively not enough categories in the above dataset (compared to 100
| Backbone | Type | Download | Configs | | Backbone | Type | Download | Configs |
| :---------------| :---------------| :---------------| :--------------- | :---------------| :---------------| :---------------| :---------------
| ResNet50-vd-FPN-Dcnv2 | Cascade Faster | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.yml) | | ResNet50-vd-FPN-Dcnv2 | Cascade Faster | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r50_vd_fpn_gen_server_side.yml) |
| ResNet101-vd-FPN-Dcnv2 | Cascade Faster | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.yml) | | ResNet101-vd-FPN-Dcnv2 | Cascade Faster | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/generic/cascade_rcnn_dcn_r101_vd_fpn_gen_server_side.yml) |
| CBResNet101-vd-FPN | Cascade Faster | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_cbr101_vd_fpn_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/rcnn_enhance/generic/cascade_rcnn_cbr101_vd_fpn_server_side.yml) | | CBResNet101-vd-FPN | Cascade Faster | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_cbr101_vd_fpn_server_side.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/rcnn_enhance/generic/cascade_rcnn_cbr101_vd_fpn_server_side.yml) |
static/docs/featured_model/YOLOv3_ENHANCEMENT.md
浏览文件 @ bf79506a
...@@ -43,12 +43,12 @@ python tools/train.py -c configs/dcn/yolov3_r50vd_dcn_db_iouloss_obj365_pretrain ...@@ -43,12 +43,12 @@ python tools/train.py -c configs/dcn/yolov3_r50vd_dcn_db_iouloss_obj365_pretrain
| 模型 | 预训练模型 | 验证集 mAP | V100 python 预测速度（FPS）<sup>[1](#1)</sup> | V100 paddle预测库速度（ms/image）<sup>[2](#2)</sup> | P4 paddle预测库速度(ms/image) <sup>[2](#2)</sup> | 下载 | 配置文件 | | 模型 | 预训练模型 | 验证集 mAP | V100 python 预测速度（FPS）<sup>[1](#1)</sup> | V100 paddle预测库速度（ms/image）<sup>[2](#2)</sup> | P4 paddle预测库速度(ms/image) <sup>[2](#2)</sup> | 下载 | 配置文件 |
| :--------------------------------------: | :----------------------------------------------------------: | :--------: | :--------: | :------------------------------------: | :----------------------------------------------------------: | :--------: | :--------: | | :--------------------------------------: | :----------------------------------------------------------: | :--------: | :--------: | :------------------------------------: | :----------------------------------------------------------: | :--------: | :--------: |
| YOLOv3 DarkNet | [DarkNet pretrain](https://paddle-imagenet-models-name.bj.bcebos.com/DarkNet53_pretrained.tar) | 38.9 | 48.55 | 原生：19.63<br>tensorRT-FP32: 18.01<br>tensorRT-FP16: 11.47 | 原生：54.10<br>tensorRT-FP32: 47.44 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet.yml) | | YOLOv3 DarkNet | [DarkNet pretrain](https://paddle-imagenet-models-name.bj.bcebos.com/DarkNet53_pretrained.tar) | 38.9 | 48.55 | 原生：19.63<br>tensorRT-FP32: 18.01<br>tensorRT-FP16: 11.47 | 原生：54.10<br>tensorRT-FP32: 47.44 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/yolov3_darknet.yml) |
| YOLOv3 ResNet50_vd DCN | [ImageNet pretrain](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar) | 39.1 | 50.80 | 原生：17.04<br>tensorRT-FP32: 16.28<br>tensorRT-FP16: 11.16 | 原生：40.01<br>tensorRT-FP32: 36.66 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_imagenet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn/yolov3_r50vd_dcn.yml) | | YOLOv3 ResNet50_vd DCN | [ImageNet pretrain](https://paddle-imagenet-models-name.bj.bcebos.com/ResNet50_vd_pretrained.tar) | 39.1 | 50.80 | 原生：17.04<br>tensorRT-FP32: 16.28<br>tensorRT-FP16: 11.16 | 原生：40.01<br>tensorRT-FP32: 36.66 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_imagenet.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn/yolov3_r50vd_dcn.yml) |
| YOLOv3 ResNet50_vd DCN | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 42.5 | 50.41 | 原生：16.76<br>tensorRT-FP32: 16.04<br>tensorRT-FP16: 10.70 | 原生：39.64<br>tensorRT-FP32: 35.93 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_obj365_v2.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn/yolov3_r50vd_dcn_obj365_pretrained_coco.yml) | | YOLOv3 ResNet50_vd DCN | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 42.5 | 50.41 | 原生：16.76<br>tensorRT-FP32: 16.04<br>tensorRT-FP16: 10.70 | 原生：39.64<br>tensorRT-FP32: 35.93 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_obj365_v2.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn/yolov3_r50vd_dcn_obj365_pretrained_coco.yml) |
| YOLOv3 ResNet50_vd DCN DropBlock | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 42.8 | 49.97 | 原生：16.55<br>tensorRT-FP32: 16.07<br>tensorRT-FP16: 10.69 | 原生：39.72<br/>tensorRT-FP32: 35.98 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_obj365_dropblock.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn/yolov3_r50vd_dcn_db_obj365_pretrained_coco.yml) | | YOLOv3 ResNet50_vd DCN DropBlock | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 42.8 | 49.97 | 原生：16.55<br>tensorRT-FP32: 16.07<br>tensorRT-FP16: 10.69 | 原生：39.72<br/>tensorRT-FP32: 35.98 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_obj365_dropblock.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn/yolov3_r50vd_dcn_db_obj365_pretrained_coco.yml) |
| YOLOv3 ResNet50_vd DCN DropBlock IoULoss | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 43.2 | 49.91 | 原生：16.46<br>tensorRT-FP32: 15.83<br>tensorRT-FP16: 10.80 | 原生：39.58<br/>tensorRT-FP32: 35.61 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_obj365_dropblock_iouloss.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn/yolov3_r50vd_dcn_db_iouloss_obj365_pretrained_coco.yml) | | YOLOv3 ResNet50_vd DCN DropBlock IoULoss | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 43.2 | 49.91 | 原生：16.46<br>tensorRT-FP32: 15.83<br>tensorRT-FP16: 10.80 | 原生：39.58<br/>tensorRT-FP32: 35.61 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_obj365_dropblock_iouloss.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn/yolov3_r50vd_dcn_db_iouloss_obj365_pretrained_coco.yml) |
| YOLOv3 ResNet50_vd DCN DropBlock IoU-Aware | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 43.6 | 48.19 | 原生：17.74<br>tensorRT-FP32: 16.73<br>tensorRT-FP16: 11.74 | 原生：41.39<br/>tensorRT-FP32: 37.75 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_db_iouaware_obj365_pretrained_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn/yolov3_r50vd_dcn_db_iouaware_obj365_pretrained_coco.yml) | | YOLOv3 ResNet50_vd DCN DropBlock IoU-Aware | [Object365 pretrain](https://paddlemodels.bj.bcebos.com/object_detection/ResNet50_vd_dcn_db_obj365_pretrained.tar) | 43.6 | 48.19 | 原生：17.74<br>tensorRT-FP32: 16.73<br>tensorRT-FP16: 11.74 | 原生：41.39<br/>tensorRT-FP32: 37.75 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_r50vd_dcn_db_iouaware_obj365_pretrained_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn/yolov3_r50vd_dcn_db_iouaware_obj365_pretrained_coco.yml) |
<a name="1">[1]</a>V100 python 预测速度是在一张Tesla V100的GPU上通过```tools/eval.py```测试所有验证集得到，单位是fps(图片数/秒), cuDNN版本是7.5，包括数据加载、网络前向执行和后处理, batch size是1。 <a name="1">[1]</a>V100 python 预测速度是在一张Tesla V100的GPU上通过```tools/eval.py```测试所有验证集得到，单位是fps(图片数/秒), cuDNN版本是7.5，包括数据加载、网络前向执行和后处理, batch size是1。
......
static/docs/featured_model/champion_model/CACascadeRCNN.md
浏览文件 @ bf79506a
...@@ -38,7 +38,7 @@ python tools/train.py -c configs/obj365/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas.y ...@@ -38,7 +38,7 @@ python tools/train.py -c configs/obj365/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas.y
| 模型 | 验证集 mAP | 下载链接 | 配置文件 | | 模型 | 验证集 mAP | 下载链接 | 配置文件 |
| :-----------------: | :--------: | :----------------------------------------------------------: | :--------: | | :-----------------: | :--------: | :----------------------------------------------------------: | :--------: |
| CACascadeRCNN SE154 | 31.6 | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas_obj365.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/obj365/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas.yml) | | CACascadeRCNN SE154 | 31.6 | [model](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas_obj365.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/obj365/cascade_rcnn_dcnv2_se154_vd_fpn_gn_cas.yml) |
## 模型效果 ## 模型效果
......
static/docs/featured_model/champion_model/OIDV5_BASELINE_MODEL.md
浏览文件 @ bf79506a
...@@ -19,15 +19,15 @@ OIDV5模型训练结果如下。 ...@@ -19,15 +19,15 @@ OIDV5模型训练结果如下。
| 模型结构 | Public/Private Score | 下载链接 | 配置文件 | | 模型结构 | Public/Private Score | 下载链接 | 配置文件 |
| :-----------------: | :--------: | :----------------------------------------------------------: | :--------: | | :-----------------: | :--------: | :----------------------------------------------------------: | :--------: |
| CascadeCARCNN-FPN-Dcnv2-Nonlocal ResNet200-vd | 0.62690/0.59459 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/oidv5_cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/oidv5/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml) | | CascadeCARCNN-FPN-Dcnv2-Nonlocal ResNet200-vd | 0.62690/0.59459 | [模型](https://paddlemodels.bj.bcebos.com/object_detection/oidv5_cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/oidv5/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml) |
此外，为验证模型的性能，PaddleDetection基于该模型结构，也训练了针对COCO2017和Objects365 Dataset的模型，模型和验证集指标如下表。 此外，为验证模型的性能，PaddleDetection基于该模型结构，也训练了针对COCO2017和Objects365 Dataset的模型，模型和验证集指标如下表。
| 模型结构 | 数据集 | 验证集mAP | 下载链接 | 配置文件 | | 模型结构 | 数据集 | 验证集mAP | 下载链接 | 配置文件 |
| :-----------------: | :--------: | :--------: | :----------------------------------------------------------: | :--------: | | :-----------------: | :--------: | :--------: | :----------------------------------------------------------: | :--------: |
| CascadeCARCNN-FPN-Dcnv2-Nonlocal ResNet200-vd | COCO2017 | 51.7% | [模型](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/dcn/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml) | | CascadeCARCNN-FPN-Dcnv2-Nonlocal ResNet200-vd | COCO2017 | 51.7% | [模型](https://paddlemodels.bj.bcebos.com/object_detection/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/dcn/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml) |
| CascadeCARCNN-FPN-Dcnv2-Nonlocal ResNet200-vd | Objects365 | 34.5% | [模型](https://paddlemodels.bj.bcebos.com/object_detection/obj365_cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/obj365/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml) | | CascadeCARCNN-FPN-Dcnv2-Nonlocal ResNet200-vd | Objects365 | 34.5% | [模型](https://paddlemodels.bj.bcebos.com/object_detection/obj365_cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs/obj365/cascade_rcnn_cls_aware_r200_vd_fpn_dcnv2_nonlocal_softnms.yml) |
COCO和Objects365 Dataset数据格式相同，目前只支持预测和评估。 COCO和Objects365 Dataset数据格式相同，目前只支持预测和评估。
......
static/docs/tutorials/INSTALL.md
浏览文件 @ bf79506a
...@@ -106,7 +106,7 @@ git clone https://github.com/PaddlePaddle/PaddleDetection.git ...@@ -106,7 +106,7 @@ git clone https://github.com/PaddlePaddle/PaddleDetection.git
**Install Python dependencies:** **Install Python dependencies:**
Required python packages are specified in [requirements.txt](https://github.com/PaddlePaddle/PaddleDetection/blob/master/requirements.txt), and can be installed with: Required python packages are specified in [requirements.txt](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/requirements.txt), and can be installed with:
``` ```
pip install -r requirements.txt pip install -r requirements.txt
......
static/slim/README.md
浏览文件 @ bf79506a
...@@ -81,7 +81,7 @@ ...@@ -81,7 +81,7 @@
### 蒸馏通道剪裁模型 ### 蒸馏通道剪裁模型
可通过高精度模型蒸馏通道剪裁后模型的方式，训练方法及相关示例见[蒸馏通道剪裁模型](https://github.com/PaddlePaddle/PaddleDetection/blob/master/slim/extensions/distill_pruned_model/distill_pruned_model_demo.ipynb)。 可通过高精度模型蒸馏通道剪裁后模型的方式，训练方法及相关示例见[蒸馏通道剪裁模型](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/slim/extensions/distill_pruned_model/distill_pruned_model_demo.ipynb)。
COCO数据集上蒸馏通道剪裁模型库如下。 COCO数据集上蒸馏通道剪裁模型库如下。
......
static/slim/distillation/README.md
浏览文件 @ bf79506a
...@@ -10,7 +10,7 @@ ...@@ -10,7 +10,7 @@
- [检测库的常规训练方法](https://github.com/PaddlePaddle/PaddleDetection) - [检测库的常规训练方法](https://github.com/PaddlePaddle/PaddleDetection)
- [PaddleSlim蒸馏API文档](https://paddlepaddle.github.io/PaddleSlim/api/single_distiller_api/) - [PaddleSlim蒸馏API文档](https://paddlepaddle.github.io/PaddleSlim/api/single_distiller_api/)
已发布蒸馏模型见[压缩模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/master/slim/README.md) 已发布蒸馏模型见[压缩模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/slim/README.md)
## 安装PaddleSlim ## 安装PaddleSlim
可按照[PaddleSlim使用文档](https://paddlepaddle.github.io/PaddleSlim/)中的步骤安装PaddleSlim 可按照[PaddleSlim使用文档](https://paddlepaddle.github.io/PaddleSlim/)中的步骤安装PaddleSlim
...@@ -89,7 +89,7 @@ yolo_output_names = [ ...@@ -89,7 +89,7 @@ yolo_output_names = [
## 训练 ## 训练
根据[PaddleDetection/tools/train.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/train.py)编写压缩脚本`distill.py`。 根据[PaddleDetection/tools/train.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/train.py)编写压缩脚本`distill.py`。
在该脚本中定义了teacher_model和student_model，用teacher_model的输出指导student_model的训练 在该脚本中定义了teacher_model和student_model，用teacher_model的输出指导student_model的训练
### 执行示例 ### 执行示例
...@@ -178,7 +178,7 @@ python -u slim/distillation/distill.py \ ...@@ -178,7 +178,7 @@ python -u slim/distillation/distill.py \
## 评估 ## 评估
每隔`snap_shot_iter`步后会保存一个checkpoint模型可以用于评估，使用PaddleDetection目录下[tools/eval.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/eval.py)评估脚本，并指定`weights`为训练得到的模型路径 每隔`snap_shot_iter`步后会保存一个checkpoint模型可以用于评估，使用PaddleDetection目录下[tools/eval.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/eval.py)评估脚本，并指定`weights`为训练得到的模型路径
运行命令为： 运行命令为：
```bash ```bash
...@@ -197,7 +197,7 @@ python -u tools/eval.py -c configs/yolov3_mobilenet_v1.yml \ ...@@ -197,7 +197,7 @@ python -u tools/eval.py -c configs/yolov3_mobilenet_v1.yml \
## 预测 ## 预测
每隔`snap_shot_iter`步后保存的checkpoint模型也可以用于预测，使用PaddleDetection目录下[tools/infer.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/infer.py)评估脚本，并指定`weights`为训练得到的模型路径 每隔`snap_shot_iter`步后保存的checkpoint模型也可以用于预测，使用PaddleDetection目录下[tools/infer.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/infer.py)评估脚本，并指定`weights`为训练得到的模型路径
### Python预测 ### Python预测
......
static/slim/nas/README.md
浏览文件 @ bf79506a
...@@ -41,7 +41,7 @@ double_blaze_filters = [ ...@@ -41,7 +41,7 @@ double_blaze_filters = [
[self.double_filter_num[tokens[2]], self.mid_filter_num[tokens[6]], self.double_filter_num[tokens[3]], 2], [self.double_filter_num[tokens[2]], self.mid_filter_num[tokens[6]], self.double_filter_num[tokens[3]], 2],
[self.double_filter_num[tokens[3]], self.mid_filter_num[tokens[7]], self.double_filter_num[tokens[3]]]] [self.double_filter_num[tokens[3]], self.mid_filter_num[tokens[7]], self.double_filter_num[tokens[3]]]]
``` ```
blaze_filters与double_blaze_filters字段请参考[blazenet.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/ppdet/modeling/backbones/blazenet.py)中定义。 blaze_filters与double_blaze_filters字段请参考[blazenet.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/ppdet/modeling/backbones/blazenet.py)中定义。
初始化tokens为：[2, 1, 3, 8, 2, 1, 2, 1, 1]。 初始化tokens为：[2, 1, 3, 8, 2, 1, 2, 1, 1]。
...@@ -84,10 +84,10 @@ BlazeNet: ...@@ -84,10 +84,10 @@ BlazeNet:
- （2）训练、评估与预测： - （2）训练、评估与预测：
启动完整的训练评估实验，可参考PaddleDetection的[训练、评估与预测流程](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/tutorials/GETTING_STARTED_cn.md) 启动完整的训练评估实验，可参考PaddleDetection的[训练、评估与预测流程](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/tutorials/GETTING_STARTED_cn.md)
## 实验结果 ## 实验结果
请参考[人脸检测模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/featured_model/FACE_DETECTION.md)中BlazeFace-NAS的实验结果。 请参考[人脸检测模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/featured_model/FACE_DETECTION.md)中BlazeFace-NAS的实验结果。
## FAQ ## FAQ
- 运行报错：`socket.error: [Errno 98] Address already in use`。 - 运行报错：`socket.error: [Errno 98] Address already in use`。
......
static/slim/prune/README.md
浏览文件 @ bf79506a
# 卷积层通道剪裁教程 # 卷积层通道剪裁教程
请确保已正确[安装PaddleDetection](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/tutorials/INSTALL_cn.md)及其依赖。 请确保已正确[安装PaddleDetection](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/tutorials/INSTALL_cn.md)及其依赖。
该文档介绍如何使用[PaddleSlim](https://paddlepaddle.github.io/PaddleSlim)的卷积通道剪裁接口对检测库中的模型的卷积层的通道数进行剪裁。 该文档介绍如何使用[PaddleSlim](https://paddlepaddle.github.io/PaddleSlim)的卷积通道剪裁接口对检测库中的模型的卷积层的通道数进行剪裁。
...@@ -8,15 +8,15 @@ ...@@ -8,15 +8,15 @@
该教程中所示操作，如无特殊说明，均在`PaddleDetection/`路径下执行。 该教程中所示操作，如无特殊说明，均在`PaddleDetection/`路径下执行。
已发布裁剪模型见[压缩模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/master/slim/README.md) 已发布裁剪模型见[压缩模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/slim/README.md)
## 1. 数据准备 ## 1. 数据准备
请参考检测库[数据下载](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/tutorials/INSTALL_cn.md)文档准备数据。 请参考检测库[数据下载](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/tutorials/INSTALL_cn.md)文档准备数据。
## 2. 模型选择 ## 2. 模型选择
通过`-c`选项指定待裁剪模型的配置文件的相对路径，更多可选配置文件请参考: [检测库配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs) 通过`-c`选项指定待裁剪模型的配置文件的相对路径，更多可选配置文件请参考: [检测库配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/develop/static/configs)
对于剪裁任务，原模型的权重不一定对剪裁后的模型训练的重训练有贡献，所以加载原模型的权重不是必需的步骤。 对于剪裁任务，原模型的权重不一定对剪裁后的模型训练的重训练有贡献，所以加载原模型的权重不是必需的步骤。
...@@ -32,7 +32,7 @@ ...@@ -32,7 +32,7 @@
-o weights=output/yolov3_mobilenet_v1_voc/model_final -o weights=output/yolov3_mobilenet_v1_voc/model_final
``` ```
官方已发布的模型请参考: [模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/MODEL_ZOO_cn.md) 官方已发布的模型请参考: [模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/MODEL_ZOO_cn.md)
## 3. 确定待分析参数 ## 3. 确定待分析参数
...@@ -49,7 +49,7 @@ python slim/prune/prune.py \ ...@@ -49,7 +49,7 @@ python slim/prune/prune.py \
## 4. 分析待剪裁参数敏感度 ## 4. 分析待剪裁参数敏感度
可通过敏感度分析脚本分析待剪裁参数敏感度得到合适的剪裁率，敏感度分析工具见[敏感度分析](https://github.com/PaddlePaddle/PaddleDetection/blob/master/slim/sensitive/README.md)。 可通过敏感度分析脚本分析待剪裁参数敏感度得到合适的剪裁率，敏感度分析工具见[敏感度分析](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/slim/sensitive/README.md)。
## 5. 启动剪裁任务 ## 5. 启动剪裁任务
...@@ -97,16 +97,16 @@ python slim/prune/export_model.py \ ...@@ -97,16 +97,16 @@ python slim/prune/export_model.py \
**当前PaddleSlim的剪裁功能不支持剪裁循环体或条件判断语句块内的卷积层，请避免剪裁循环和判断语句块前的一个卷积和语句块内部的卷积。** **当前PaddleSlim的剪裁功能不支持剪裁循环体或条件判断语句块内的卷积层，请避免剪裁循环和判断语句块前的一个卷积和语句块内部的卷积。**
对于[faster_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/faster_rcnn_r50_1x.yml)或[mask_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/mask_rcnn_r50_1x.yml)网络，请剪裁卷积`res4f_branch2c`之前的卷积。 对于[faster_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/faster_rcnn_r50_1x.yml)或[mask_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/mask_rcnn_r50_1x.yml)网络，请剪裁卷积`res4f_branch2c`之前的卷积。
对[faster_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/faster_rcnn_r50_1x.yml)剪裁示例如下： 对[faster_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/faster_rcnn_r50_1x.yml)剪裁示例如下：
``` ```
# demo for faster_rcnn_r50 # demo for faster_rcnn_r50
python slim/prune/prune.py -c ./configs/faster_rcnn_r50_1x.yml --pruned_params "res4f_branch2b_weights,res4f_branch2a_weights" --pruned_ratios="0.3,0.4" --eval python slim/prune/prune.py -c ./configs/faster_rcnn_r50_1x.yml --pruned_params "res4f_branch2b_weights,res4f_branch2a_weights" --pruned_ratios="0.3,0.4" --eval
``` ```
对[mask_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs/mask_rcnn_r50_1x.yml)剪裁示例如下： 对[mask_rcnn_r50](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs/mask_rcnn_r50_1x.yml)剪裁示例如下：
``` ```
# demo for mask_rcnn_r50 # demo for mask_rcnn_r50
......
static/slim/quantization/README.md
浏览文件 @ bf79506a
...@@ -19,7 +19,7 @@ ...@@ -19,7 +19,7 @@
## 训练 ## 训练
根据 [tools/train.py](https://github.com/PaddlePaddle/PaddleDetection/blob/master/tools/train.py) 编写压缩脚本train.py。脚本中量化的步骤如下。 根据 [tools/train.py](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/tools/train.py) 编写压缩脚本train.py。脚本中量化的步骤如下。
### 定义量化配置 ### 定义量化配置
config = { config = {
......
static/slim/sensitive/README.md
浏览文件 @ bf79506a
# 卷积层敏感度分析教程 # 卷积层敏感度分析教程
请确保已正确[安装PaddleDetection](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/tutorials/INSTALL_cn.md)及其依赖。 请确保已正确[安装PaddleDetection](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/tutorials/INSTALL_cn.md)及其依赖。
该文档介绍如何使用[PaddleSlim](https://paddlepaddle.github.io/PaddleSlim)的敏感度分析接口对检测库中的模型的卷积层进行敏感度分析。 该文档介绍如何使用[PaddleSlim](https://paddlepaddle.github.io/PaddleSlim)的敏感度分析接口对检测库中的模型的卷积层进行敏感度分析。
...@@ -10,11 +10,11 @@ ...@@ -10,11 +10,11 @@
## 数据准备 ## 数据准备
请参考检测库[数据模块](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/tutorials/INSTALL_cn.md)文档准备数据。 请参考检测库[数据模块](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/tutorials/INSTALL_cn.md)文档准备数据。
## 模型选择 ## 模型选择
通过`-c`选项指定待分析模型的配置文件的相对路径，更多可选配置文件请参考: [检测库配置文件](https://github.com/PaddlePaddle/PaddleDetection/blob/master/configs) 通过`-c`选项指定待分析模型的配置文件的相对路径，更多可选配置文件请参考: [检测库配置文件](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/configs)
通过`-o weights`指定模型的权重，可以指定url或本地文件系统的路径。如下所示： 通过`-o weights`指定模型的权重，可以指定url或本地文件系统的路径。如下所示：
...@@ -28,7 +28,7 @@ ...@@ -28,7 +28,7 @@
-o weights=output/yolov3_mobilenet_v1_voc/model_final -o weights=output/yolov3_mobilenet_v1_voc/model_final
``` ```
官方已发布的模型请参考: [模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/master/docs/MODEL_ZOO_cn.md) 官方已发布的模型请参考: [模型库](https://github.com/PaddlePaddle/PaddleDetection/blob/develop/static/docs/MODEL_ZOO_cn.md)
## 确定待分析参数 ## 确定待分析参数
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或