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 c3aad6b6 编写于 作者: G Guanghua Yu 提交者： GitHub
浏览文件
操作

add yolov3_darknet diou_loss model (#1041)

上级 15dd0fab
隐藏空白更改
内联 并排
Showing with 6 addition and 2 deletion
docs/MODEL_ZOO.md
浏览文件 @ c3aad6b6
...@@ -149,11 +149,12 @@ The backbone models pretrained on ImageNet are available. All backbone models ar ...@@ -149,11 +149,12 @@ The backbone models pretrained on ImageNet are available. All backbone models ar
### YOLO v3 on Pascal VOC ### YOLO v3 on Pascal VOC
| Backbone | Size | Image/gpu | Lr schd | Inf time (fps) | Box AP | Download | Configs | | Backbone | Size | Image/gpu | Lr schd | Inf time (fps) | Box AP(0.5) | Download | Configs |
| :----------- | :--: | :-------: | :-----: | :------------: | :----: | :----------------------------------------------------------: | :----: | | :----------- | :--: | :-------: | :-----: | :------------: | :----: | :----------------------------------------------------------: | :----: |
| DarkNet53 | 608 | 8 | 270e | 54.977 | 83.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) | | DarkNet53 | 608 | 8 | 270e | 54.977 | 83.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) |
| DarkNet53 | 416 | 8 | 270e | - | 83.6 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) | | DarkNet53 | 416 | 8 | 270e | - | 83.6 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) |
| DarkNet53 | 320 | 8 | 270e | - | 82.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) | | DarkNet53 | 320 | 8 | 270e | - | 82.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) |
| DarkNet53 Diou-Loss | 608 | 8 | 270e | - | 83.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc_diouloss.pdparams) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc_diouloss.yml) |
| MobileNet-V1 | 608 | 8 | 270e | 104.291 | 76.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) | | MobileNet-V1 | 608 | 8 | 270e | 104.291 | 76.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) |
| MobileNet-V1 | 416 | 8 | 270e | - | 76.7 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) | | MobileNet-V1 | 416 | 8 | 270e | - | 76.7 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) |
| MobileNet-V1 | 320 | 8 | 270e | - | 75.3 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) | | MobileNet-V1 | 320 | 8 | 270e | - | 75.3 | [model](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) |
...@@ -169,6 +170,7 @@ improved performance mainly by using L1 loss in bounding box width and height re ...@@ -169,6 +170,7 @@ improved performance mainly by using L1 loss in bounding box width and height re
randomly color distortion, randomly cropping, randomly expansion, randomly interpolation method, randomly flippling. YOLO v3 used randomly randomly color distortion, randomly cropping, randomly expansion, randomly interpolation method, randomly flippling. YOLO v3 used randomly
reshaped minibatch in training, inferences can be performed on different image sizes with the same model weights, and we provided evaluation reshaped minibatch in training, inferences can be performed on different image sizes with the same model weights, and we provided evaluation
results of image size 608/416/320 above. Deformable conv is added on stage 5 of backbone. results of image size 608/416/320 above. Deformable conv is added on stage 5 of backbone.
- Compared with YOLOv3-DarkNet53, the average AP of YOLOv3-DarkNet53 with Diou-Loss increases about 2% in VOC dataset.
- YOLO v3 enhanced model improves the precision to 43.6 involved with deformable conv, dropblock, IoU loss and IoU aware. See more details in [YOLOv3_ENHANCEMENT](./featured_model/YOLOv3_ENHANCEMENT.md) - YOLO v3 enhanced model improves the precision to 43.6 involved with deformable conv, dropblock, IoU loss and IoU aware. See more details in [YOLOv3_ENHANCEMENT](./featured_model/YOLOv3_ENHANCEMENT.md)
### RetinaNet ### RetinaNet
...@@ -212,7 +214,7 @@ results of image size 608/416/320 above. Deformable conv is added on stage 5 of ...@@ -212,7 +214,7 @@ results of image size 608/416/320 above. Deformable conv is added on stage 5 of
### SSD on Pascal VOC ### SSD on Pascal VOC
| Backbone | Size | Image/gpu | Lr schd | Inf time (fps) | Box AP | Download | Configs | | Backbone | Size | Image/gpu | Lr schd | Inf time (fps) | Box AP(0.5) | Download | Configs |
| :----------- | :--: | :-------: | :-----: | :------------: | :----: | :----------------------------------------------------------: | :----: | | :----------- | :--: | :-------: | :-----: | :------------: | :----: | :----------------------------------------------------------: | :----: |
| MobileNet v1 | 300 | 32 | 120e | 159.543 | 73.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/ssd_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/ssd/ssd_mobilenet_v1_voc.yml) | | MobileNet v1 | 300 | 32 | 120e | 159.543 | 73.2 | [model](https://paddlemodels.bj.bcebos.com/object_detection/ssd_mobilenet_v1_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/ssd/ssd_mobilenet_v1_voc.yml) |
| VGG16 | 300 | 8 | 240e | 117.279 | 77.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/ssd_vgg16_300_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/ssd/ssd_vgg16_300_voc.yml) | | VGG16 | 300 | 8 | 240e | 117.279 | 77.5 | [model](https://paddlemodels.bj.bcebos.com/object_detection/ssd_vgg16_300_voc.tar) | [config](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/ssd/ssd_vgg16_300_voc.yml) |
......
docs/MODEL_ZOO_cn.md
浏览文件 @ c3aad6b6
...@@ -150,6 +150,7 @@ Paddle提供基于ImageNet的骨架网络预训练模型。所有预训练模型 ...@@ -150,6 +150,7 @@ Paddle提供基于ImageNet的骨架网络预训练模型。所有预训练模型
| DarkNet53 | 608 | 8 | 270e | 54.977 | 83.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) | | DarkNet53 | 608 | 8 | 270e | 54.977 | 83.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) |
| DarkNet53 | 416 | 8 | 270e | - | 83.6 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) | | DarkNet53 | 416 | 8 | 270e | - | 83.6 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) |
| DarkNet53 | 320 | 8 | 270e | - | 82.2 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) | | DarkNet53 | 320 | 8 | 270e | - | 82.2 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc.yml) |
| DarkNet53 Diou-Loss | 608 | 8 | 270e | - | 83.5 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_darknet_voc_diouloss.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_darknet_voc_diouloss.yml) |
| MobileNet-V1 | 608 | 8 | 270e | 104.291 | 76.2 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) | | MobileNet-V1 | 608 | 8 | 270e | 104.291 | 76.2 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) |
| MobileNet-V1 | 416 | 8 | 270e | - | 76.7 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) | | MobileNet-V1 | 416 | 8 | 270e | - | 76.7 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) |
| MobileNet-V1 | 320 | 8 | 270e | - | 75.3 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) | | MobileNet-V1 | 320 | 8 | 270e | - | 75.3 | [下载链接](https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/configs/yolov3_mobilenet_v1_voc.yml) |
...@@ -161,6 +162,7 @@ Paddle提供基于ImageNet的骨架网络预训练模型。所有预训练模型 ...@@ -161,6 +162,7 @@ Paddle提供基于ImageNet的骨架网络预训练模型。所有预训练模型
- 上表中也提供了原论文[YOLOv3](https://arxiv.org/abs/1804.02767)中YOLOv3-DarkNet53的精度，我们的实现版本主要从在bounding box的宽度和高度回归上使用了L1损失，图像mixup和label smooth等方法优化了其精度。 - 上表中也提供了原论文[YOLOv3](https://arxiv.org/abs/1804.02767)中YOLOv3-DarkNet53的精度，我们的实现版本主要从在bounding box的宽度和高度回归上使用了L1损失，图像mixup和label smooth等方法优化了其精度。
- YOLO v3在8卡，总batch size为64下训练270轮。数据增强包括：mixup, 随机颜色失真，随机剪裁，随机扩张，随机插值法，随机翻转。YOLO v3在训练阶段对minibatch采用随机reshape，可以采用相同的模型测试不同尺寸图片，我们分别提供了尺寸为608/416/320大小的测试结果。deformable卷积作用在骨架网络5阶段。 - YOLO v3在8卡，总batch size为64下训练270轮。数据增强包括：mixup, 随机颜色失真，随机剪裁，随机扩张，随机插值法，随机翻转。YOLO v3在训练阶段对minibatch采用随机reshape，可以采用相同的模型测试不同尺寸图片，我们分别提供了尺寸为608/416/320大小的测试结果。deformable卷积作用在骨架网络5阶段。
- 在YOLOv3-DarkNet53模型基础上使用Diou-Loss后，在VOC数据集上该模型平均mAP比原模型高大约2%。
- YOLO v3增强版模型通过引入可变形卷积，dropblock，IoU loss和Iou aware，将精度进一步提升至43.6， 详情见[YOLOv3增强模型](./featured_model/YOLOv3_ENHANCEMENT.md) - YOLO v3增强版模型通过引入可变形卷积，dropblock，IoU loss和Iou aware，将精度进一步提升至43.6， 详情见[YOLOv3增强模型](./featured_model/YOLOv3_ENHANCEMENT.md)
### RetinaNet ### RetinaNet
......
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