未验证 提交 1a66a9d2 编写于 作者: K Kaipeng Deng 提交者: GitHub

add 2.0rc changelog (#2254)

上级 7c3def89
...@@ -250,7 +250,7 @@ PaddleDetection模块化地实现了多种主流目标检测算法,提供了 ...@@ -250,7 +250,7 @@ PaddleDetection模块化地实现了多种主流目标检测算法,提供了
- [使用SSD-MobileNetv1完成一个项目--准备数据集到完成树莓派部署](https://github.com/PaddleCV-FAQ/PaddleDetection-FAQ/blob/main/Lite%E9%83%A8%E7%BD%B2/ssd_mobilenet_v1_for_raspi.md) - [使用SSD-MobileNetv1完成一个项目--准备数据集到完成树莓派部署](https://github.com/PaddleCV-FAQ/PaddleDetection-FAQ/blob/main/Lite%E9%83%A8%E7%BD%B2/ssd_mobilenet_v1_for_raspi.md)
## 版本更新 ## 版本更新
v0.5.0版本已经在`11/2020`发布,增加SOLOv2等多个模型,新增Android移动端Demo,新增PACT量化策略,新增目标检测全流程和Jetson平台部署等教程,提升易用性,修复已知诸多bug等,详细内容请参考[版本更新文档](docs/CHANGELOG.md) v2.0-rc版本已经在`02/2021`发布,新增动态图版本,支持RCNN, YOLOv3, PP-YOLO, SSD/SSDLite, FCOS, TTFNet, SOLOv2等系列模型,支持模型剪裁和量化,支持预测部署及TensorRT推理加速,详细内容请参考[版本更新文档](docs/CHANGELOG.md)
## 许可证书 ## 许可证书
本项目的发布受[Apache 2.0 license](LICENSE)许可认证。 本项目的发布受[Apache 2.0 license](LICENSE)许可认证。
......
...@@ -260,7 +260,7 @@ All these models can be get in [Model Zoo](#ModelZoo) ...@@ -260,7 +260,7 @@ All these models can be get in [Model Zoo](#ModelZoo)
## Updates ## Updates
v0.5.0 was released at `11/2020`, add SOLOv2, etc. And add Android mobile demo, add PACT quantization strategy, add object detection full-process documentation and Jetson platform deployment tutorials, and improved ease of use, fix many known bugs, etc. Please refer to [change log](docs/CHANGELOG.md) for details. v2.0-rc was released at `02/2021`, add dygraph version, which supports RCNN, YOLOv3, PP-YOLO, SSD/SSDLite, FCOS, TTFNet, SOLOv2, etc. supports model pruning and quantization, supports deploying and accelerating by TensorRT, etc. Please refer to [change log](docs/CHANGELOG.md) for details.
## License ## License
......
...@@ -2,6 +2,31 @@ ...@@ -2,6 +2,31 @@
## 最新版本信息 ## 最新版本信息
### 2.0-rc(02.23/2021)
- 动态图模型丰富度提升:
- 优化RCNN模型组网及训练方式,RCNN系列模型精度提升(依赖Paddle develop或2.0.1版本)
- 新增支持SSDLite,FCOS,TTFNet,SOLOv2系列模型
- 新增行人和车辆垂类目标检测模型
- 新增动态图基础模块:
- 新增MobileNetV3,HRNet骨干网络
- 优化RoIAlign计算逻辑,RCNN系列模型精度提升(依赖Paddle develop或2.0.1版本)
- 新增支持Synchronized Batch Norm
- 新增支持Modulated Deformable Convolution
- 预测部署:
- 发布动态图python、C++、Serving部署解决方案及文档,支持Faster RCNN,Mask RCNN,YOLOv3,PP-YOLO,SSD,TTFNet,FCOS,SOLOv2等系列模型预测部署
- 动态图预测部署支持TensorRT模式FP32,FP16推理加速
- 检测模型压缩:
- 裁剪:新增动态图裁剪支持,并发布YOLOv3-MobileNetV1裁剪模型
- 量化:新增动态图量化支持,并发布YOLOv3-MobileNetV1和YOLOv3-MobileNetV3量化模型
- 文档:
- 新增动态图入门教程文档:包含安装说明,快速开始,准备数据,训练/评估/预测流程文档
- 新增动态图进阶教程文档:包含模型压缩、推理部署文档
- 新增动态图模型库文档
### v2.0-beta(12.20/2020) ### v2.0-beta(12.20/2020)
- 动态图支持: - 动态图支持:
- 支持Faster-RCNN, Mask-RCNN, FPN, Cascade Faster/Mask RCNN, YOLOv3和SSD模型,试用版本。 - 支持Faster-RCNN, Mask-RCNN, FPN, Cascade Faster/Mask RCNN, YOLOv3和SSD模型,试用版本。
......
...@@ -7,6 +7,8 @@ ...@@ -7,6 +7,8 @@
| ResNet50-FPN | Cascade Faster | 1 | 1x | ---- | 41.1 | - | [下载链接](https://paddledet.bj.bcebos.com/models/cascade_rcnn_r50_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco.yml) | | ResNet50-FPN | Cascade Faster | 1 | 1x | ---- | 41.1 | - | [下载链接](https://paddledet.bj.bcebos.com/models/cascade_rcnn_r50_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/cascade_rcnn/cascade_rcnn_r50_fpn_1x_coco.yml) |
| ResNet50-FPN | Cascade Mask | 1 | 1x | ---- | 41.8 | 36.3 | [下载链接](https://paddledet.bj.bcebos.com/models/cascade_mask_rcnn_r50_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco.yml) | | ResNet50-FPN | Cascade Mask | 1 | 1x | ---- | 41.8 | 36.3 | [下载链接](https://paddledet.bj.bcebos.com/models/cascade_mask_rcnn_r50_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/cascade_rcnn/cascade_mask_rcnn_r50_fpn_1x_coco.yml) |
**注意:** Cascade R-CNN模型精度依赖Paddle develop分支修改,精度复现须使用[每日版本](https://www.paddlepaddle.org.cn/documentation/docs/zh/install/Tables.html#whl-dev)或2.0.1版本(将于2021.03发布),使用Paddle 2.0.0版本会有少量精度损失。
## Citations ## Citations
``` ```
@article{Cai_2019, @article{Cai_2019,
......
...@@ -19,6 +19,8 @@ ...@@ -19,6 +19,8 @@
| ResNeXt101-vd-FPN | Faster | 1 | 1x | ---- | 43.4 | [下载链接](https://paddledet.bj.bcebos.com/models/faster_rcnn_x101_vd_64x4d_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco.yml) | | ResNeXt101-vd-FPN | Faster | 1 | 1x | ---- | 43.4 | [下载链接](https://paddledet.bj.bcebos.com/models/faster_rcnn_x101_vd_64x4d_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_1x_coco.yml) |
| ResNeXt101-vd-FPN | Faster | 1 | 2x | ---- | 44.0 | [下载链接](https://paddledet.bj.bcebos.com/models/faster_rcnn_x101_vd_64x4d_fpn_2x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_2x_coco.yml) | | ResNeXt101-vd-FPN | Faster | 1 | 2x | ---- | 44.0 | [下载链接](https://paddledet.bj.bcebos.com/models/faster_rcnn_x101_vd_64x4d_fpn_2x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/faster_rcnn/faster_rcnn_x101_vd_64x4d_fpn_2x_coco.yml) |
**注意:** Faster R-CNN模型精度依赖Paddle develop分支修改,精度复现须使用[每日版本](https://www.paddlepaddle.org.cn/documentation/docs/zh/install/Tables.html#whl-dev)或2.0.1版本(将于2021.03发布),使用Paddle 2.0.0版本会有少量精度损失。
## Citations ## Citations
``` ```
@article{Ren_2017, @article{Ren_2017,
......
...@@ -19,6 +19,7 @@ FCOS (Fully Convolutional One-Stage Object Detection) is a fast anchor-free obje ...@@ -19,6 +19,7 @@ FCOS (Fully Convolutional One-Stage Object Detection) is a fast anchor-free obje
**Notes:** **Notes:**
- FCOS is trained on COCO train2017 dataset and evaluated on val2017 results of `mAP(IoU=0.5:0.95)`. - FCOS is trained on COCO train2017 dataset and evaluated on val2017 results of `mAP(IoU=0.5:0.95)`.
- FCOS training performace is dependented on Paddle develop branch, performance reproduction shoule based on [Paddle daily version](https://www.paddlepaddle.org.cn/documentation/docs/zh/install/Tables.html#whl-dev) or Paddle 2.0.1(will be published on 2021.03), performace will loss slightly is training base on Paddle 2.0.0
## Citations ## Citations
``` ```
......
...@@ -15,6 +15,8 @@ ...@@ -15,6 +15,8 @@
| ResNeXt101-vd-FPN | Mask | 1 | 1x | ---- | 44.0 | 39.5 | [下载链接](https://paddledet.bj.bcebos.com/models/mask_rcnn_x101_vd_64x4d_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco.yml) | | ResNeXt101-vd-FPN | Mask | 1 | 1x | ---- | 44.0 | 39.5 | [下载链接](https://paddledet.bj.bcebos.com/models/mask_rcnn_x101_vd_64x4d_fpn_1x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_1x_coco.yml) |
| ResNeXt101-vd-FPN | Mask | 1 | 2x | ---- | - | - | [下载链接](https://paddledet.bj.bcebos.com/models/mask_rcnn_x101_vd_64x4d_fpn_2x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_2x_coco.yml) | | ResNeXt101-vd-FPN | Mask | 1 | 2x | ---- | - | - | [下载链接](https://paddledet.bj.bcebos.com/models/mask_rcnn_x101_vd_64x4d_fpn_2x_coco.pdparams) | [配置文件](https://github.com/PaddlePaddle/PaddleDetection/tree/master/dygraph/configs/mask_rcnn/mask_rcnn_x101_vd_64x4d_fpn_2x_coco.yml) |
**注意:** Mask R-CNN模型精度依赖Paddle develop分支修改,精度复现须使用[每日版本](https://www.paddlepaddle.org.cn/documentation/docs/zh/install/Tables.html#whl-dev)或2.0.1版本(将于2021.03发布),使用Paddle 2.0.0版本会有少量精度损失。
## Citations ## Citations
``` ```
@article{He_2017, @article{He_2017,
......
...@@ -25,6 +25,7 @@ SOLOv2 (Segmenting Objects by Locations) is a fast instance segmentation framewo ...@@ -25,6 +25,7 @@ SOLOv2 (Segmenting Objects by Locations) is a fast instance segmentation framewo
**Notes:** **Notes:**
- SOLOv2 is trained on COCO train2017 dataset and evaluated on val2017 results of `mAP(IoU=0.5:0.95)`. - SOLOv2 is trained on COCO train2017 dataset and evaluated on val2017 results of `mAP(IoU=0.5:0.95)`.
- SOLOv2 training performace is dependented on Paddle develop branch, performance reproduction shoule based on [Paddle daily version](https://www.paddlepaddle.org.cn/documentation/docs/zh/install/Tables.html#whl-dev) or Paddle 2.0.1(will be published on 2021.03), performace will loss slightly is training base on Paddle 2.0.0
## Citations ## Citations
``` ```
......
Markdown is supported
0% .
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