mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2025-10-25 09:31:38 +08:00
1438b66c6e
* first commit for yolov7 * pybind for yolov7 * CPP README.md * CPP README.md * modified yolov7.cc * README.md * python file modify * delete license in fastdeploy/ * repush the conflict part * README.md modified * README.md modified * file path modified * file path modified * file path modified * file path modified * file path modified * README modified * README modified * move some helpers to private * add examples for yolov7 * api.md modified * api.md modified * api.md modified * YOLOv7 * yolov7 release link * yolov7 release link * yolov7 release link * copyright * change some helpers to private * change variables to const and fix documents. * gitignore * Transfer some funtions to private member of class * Transfer some funtions to private member of class * Merge from develop (#9) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> * first commit for yolor * for merge * Develop (#11) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> * Yolor (#16) * Develop (#11) (#12) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> * Develop (#13) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> * documents * documents * documents * documents * documents * documents * documents * documents * documents * documents * documents * documents * Develop (#14) * Fix compile problem in different python version (#26) * fix some usage problem in linux * Fix compile problem Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> * Add PaddleDetetion/PPYOLOE model support (#22) * add ppdet/ppyoloe * Add demo code and documents * add convert processor to vision (#27) * update .gitignore * Added checking for cmake include dir * fixed missing trt_backend option bug when init from trt * remove un-need data layout and add pre-check for dtype * changed RGB2BRG to BGR2RGB in ppcls model * add model_zoo yolov6 c++/python demo * fixed CMakeLists.txt typos * update yolov6 cpp/README.md * add yolox c++/pybind and model_zoo demo * move some helpers to private * fixed CMakeLists.txt typos * add normalize with alpha and beta * add version notes for yolov5/yolov6/yolox * add copyright to yolov5.cc * revert normalize * fixed some bugs in yolox * fixed examples/CMakeLists.txt to avoid conflicts * add convert processor to vision * format examples/CMakeLists summary * Fix bug while the inference result is empty with YOLOv5 (#29) * Add multi-label function for yolov5 * Update README.md Update doc * Update fastdeploy_runtime.cc fix variable option.trt_max_shape wrong name * Update runtime_option.md Update resnet model dynamic shape setting name from images to x * Fix bug when inference result boxes are empty * Delete detection.py Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> Co-authored-by: Jason <928090362@qq.com> * add is_dynamic for YOLO series (#22) * fix dead links Co-authored-by: Jason <jiangjiajun@baidu.com> Co-authored-by: root <root@bjyz-sys-gpu-kongming3.bjyz.baidu.com> Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com> Co-authored-by: huangjianhui <852142024@qq.com> Co-authored-by: Jason <928090362@qq.com>
69 lines
3.9 KiB
Markdown
69 lines
3.9 KiB
Markdown
# SCRFD准备部署模型
|
||
|
||
|
||
- [SCRFD](https://github.com/deepinsight/insightface/tree/17cdeab12a35efcebc2660453a8cbeae96e20950)
|
||
- （1）[官方库](https://github.com/deepinsight/insightface/)中提供的*.pt通过[导出ONNX模型](#导出ONNX模型)操作后，可进行部署；
|
||
- （2）开发者基于自己数据训练的SCRFD模型，可按照[导出ONNX模型](#%E5%AF%BC%E5%87%BAONNX%E6%A8%A1%E5%9E%8B)后，完成部署。
|
||
|
||
|
||
## 导出ONNX模型
|
||
|
||
```
|
||
#下载scrfd模型文件
|
||
e.g. download from https://onedrive.live.com/?authkey=%21ABbFJx2JMhNjhNA&id=4A83B6B633B029CC%215542&cid=4A83B6B633B029CC
|
||
|
||
# 安装官方库配置环境，此版本导出环境为：
|
||
- 手动配置环境
|
||
torch==1.8.0
|
||
mmcv==1.3.5
|
||
mmdet==2.7.0
|
||
|
||
- 通过docker配置
|
||
docker pull qyjdefdocker/onnx-scrfd-converter:v0.3
|
||
|
||
# 导出onnx格式文件
|
||
- 手动生成
|
||
python tools/scrfd2onnx.py configs/scrfd/scrfd_500m.py weights/scrfd_500m.pth --shape 640 --input-img face-xxx.jpg
|
||
|
||
- docker
|
||
docker的onnx目录中已有生成好的onnx文件
|
||
|
||
```
|
||
|
||
## 下载预训练ONNX模型
|
||
|
||
为了方便开发者的测试，下面提供了SCRFD导出的各系列模型，开发者可直接下载使用。（下表中模型的精度来源于源官方库）
|
||
| 模型 | 大小 | 精度 |
|
||
|:---------------------------------------------------------------- |:----- |:----- |
|
||
| [SCRFD-500M-kps-160](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_500m_bnkps_shape160x160.onnx) | 2.5MB | - |
|
||
| [SCRFD-500M-160](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_500m_shape160x160.onnx) | 2.2MB | - |
|
||
| [SCRFD-500M-kps-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_500m_bnkps_shape320x320.onnx) | 2.5MB | - |
|
||
| [SCRFD-500M-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_500m_shape320x320.onnx) | 2.2MB | - |
|
||
| [SCRFD-500M-kps-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_500m_bnkps_shape640x640.onnx) | 2.5MB | 90.97% |
|
||
| [SCRFD-500M-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_500m_shape640x640.onnx) | 2.2MB | 90.57% |
|
||
| [SCRFD-1G-160](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_1g_shape160x160.onnx ) | 2.5MB | - |
|
||
| [SCRFD-1G-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_1g_shape320x320.onnx) | 2.5MB | - |
|
||
| [SCRFD-1G-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_1g_shape640x640.onnx) | 2.5MB | 92.38% |
|
||
| [SCRFD-2.5G-kps-160](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_2.5g_bnkps_shape160x160.onnx) | 3.2MB | - |
|
||
| [SCRFD-2.5G-160](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_2.5g_shape160x160.onnx) | 2.6MB | - |
|
||
| [SCRFD-2.5G-kps-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_2.5g_bnkps_shape320x320.onnx) | 3.2MB | - |
|
||
| [SCRFD-2.5G-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_2.5g_shape320x320.onnx) | 2.6MB | - |
|
||
| [SCRFD-2.5G-kps-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_2.5g_bnkps_shape640x640.onnx) | 3.2MB | 93.8% |
|
||
| [SCRFD-2.5G-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_2.5g_shape640x640.onnx) | 2.6MB | 93.78% |
|
||
| [SCRFD-10G-kps-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_10g_bnkps_shape320x320.onnx) | 17MB | - |
|
||
| [SCRFD-10G-320](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_10g_shape320x320.onnx) | 15MB | - |
|
||
| [SCRFD-10G-kps-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_10g_bnkps_shape640x640.onnx) | 17MB | 95.4% |
|
||
| [SCRFD-10G-640](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_10g_shape640x640.onnx) | 15MB | 95.16% |
|
||
| [SCRFD-10G-kps-1280](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_10g_bnkps_shape1280x1280.onnx) | 17MB | - |
|
||
| [SCRFD-10G-1280](https://bj.bcebos.com/paddlehub/fastdeploy/scrfd_10g_shape1280x1280.onnx) | 15MB | - |
|
||
|
||
## 详细部署文档
|
||
|
||
- [Python部署](python)
|
||
- [C++部署](cpp)
|
||
|
||
|
||
## 版本说明
|
||
|
||
- 本版本文档和代码基于[SCRFD CommitID:17cdeab](https://github.com/deepinsight/insightface/tree/17cdeab12a35efcebc2660453a8cbeae96e20950) 编写
|