diff --git a/models/cv/detection/atss/igie/README.md b/models/cv/detection/atss/igie/README.md index c64581d413960de39d85032115e563b180e83662..413892513ae1627a9b8906f6310f4b47fa2029c8 100644 --- a/models/cv/detection/atss/igie/README.md +++ b/models/cv/detection/atss/igie/README.md @@ -24,12 +24,17 @@ pip3 install mmengine==0.10.4 ``` ### Download -Pretrained model: + +Pretrained model: Dataset: to download the validation dataset. +```bash +wget https://download.openmmlab.com/mmdetection/v2.0/atss/atss_r50_fpn_1x_coco/atss_r50_fpn_1x_coco_20200209-985f7bd0.pth +``` ### Model Conversion + ```bash # export onnx model python3 export.py --weight atss_r50_fpn_1x_coco_20200209-985f7bd0.pth --cfg atss_r50_fpn_1x_coco.py --output atss.onnx @@ -55,6 +60,6 @@ bash scripts/infer_atss_fp16_performance.sh ## Results -| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | -| :-------: | :-------: | :---------: | :-------: | :-------: | :--------: | -| ATSS | 32 | 800x800 | FP16 | 81.671 | 0.541 | \ No newline at end of file +| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | +| :---: | :-------: | :---------: | :-------: | :-----: | :--------: | +| ATSS | 32 | 800x800 | FP16 | 126.864 | 0.541 | diff --git a/models/cv/detection/fcos/igie/README.md b/models/cv/detection/fcos/igie/README.md index f10cc742410c3aaf0bd32ea27b7b177ed90f1936..2aa984220574c4ad554fd867422e6047bcb23ca0 100644 --- a/models/cv/detection/fcos/igie/README.md +++ b/models/cv/detection/fcos/igie/README.md @@ -24,12 +24,17 @@ pip3 install mmengine==0.10.4 ``` ### Download + Pretrained model: Dataset: to download the validation dataset. +```bash +wget https://download.openmmlab.com/mmdetection/v2.0/fcos/fcos_r50_caffe_fpn_gn-head_1x_coco/fcos_r50_caffe_fpn_gn-head_1x_coco-821213aa.pth +``` ### Model Conversion + ```bash # export onnx model python3 export.py --weight fcos_r50_caffe_fpn_gn-head_1x_coco-821213aa.pth --cfg fcos_r50_caffe_fpn_gn-head_1x_coco.py --output fcos.onnx @@ -55,6 +60,6 @@ bash scripts/infer_fcos_fp16_performance.sh ## Results -| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | -| :-------: | :-------: | :---------: | :-------: | :-------: | :--------: | -| FCOS | 32 | 800x800 | FP16 | 83.09 | 0.522 | \ No newline at end of file +| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | +| :---: | :-------: | :---------: | :-------: | :-----: | :--------: | +| FCOS | 32 | 800x800 | FP16 | 135.019 | 0.522 | diff --git a/models/cv/detection/fsaf/igie/README.md b/models/cv/detection/fsaf/igie/README.md index 14d56932be1695135a8f864f00eecba5c0932a83..7a587dc0e10d1c382019a0b18c7bbe0574d26994 100644 --- a/models/cv/detection/fsaf/igie/README.md +++ b/models/cv/detection/fsaf/igie/README.md @@ -24,12 +24,17 @@ pip3 install mmengine==0.10.4 ``` ### Download + Pretrained model: Dataset: to download the validation dataset. +```bash +wget https://download.openmmlab.com/mmdetection/v2.0/fsaf/fsaf_r50_fpn_1x_coco/fsaf_r50_fpn_1x_coco-94ccc51f.pth +``` ### Model Conversion + ```bash # export onnx model python3 export.py --weight fsaf_r50_fpn_1x_coco-94ccc51f.pth --cfg fsaf_r50_fpn_1x_coco.py --output fsaf.onnx @@ -55,6 +60,6 @@ bash scripts/infer_fsaf_fp16_performance.sh ## Results -| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | -| :-------: | :-------: | :---------: | :-------: | :-------: | :--------: | -| FSAF | 32 | 800x800 | FP16 | 122.35 | 0.530 | \ No newline at end of file +| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | +| :---: | :-------: | :---------: | :-------: | :-----: | :--------: | +| FSAF | 32 | 800x800 | FP16 | 178.748 | 0.530 | diff --git a/models/cv/detection/retinaface/igie/README.md b/models/cv/detection/retinaface/igie/README.md index 18da6c3c7013fba51d02d25b7e1650d9d01e1f47..a4eb4e4d406b3b225c4db9f4980a9f35b757aa3d 100755 --- a/models/cv/detection/retinaface/igie/README.md +++ b/models/cv/detection/retinaface/igie/README.md @@ -7,7 +7,14 @@ RetinaFace is an efficient single-stage face detection model that employs a mult ## Setup ### Install + ```bash +# Install libGL +## CentOS +yum install -y mesa-libGL +## Ubuntu +apt install -y libgl1-mesa-dev + pip3 install onnx pip3 install tqdm pip3 install onnxsim @@ -15,11 +22,17 @@ pip3 install opencv-python==4.6.0.66 ``` ### Download + Pretrained model: Dataset: to download the validation dataset. +```bash +wget https://github.com/biubug6/Face-Detector-1MB-with-landmark/raw/master/weights/mobilenet0.25_Final.pth +``` + ### Model Conversion + ```bash # export onnx model python3 export.py --weight mobilenet0.25_Final.pth --output retinaface.onnx @@ -35,6 +48,7 @@ export DATASETS_DIR=/Path/to/widerface/ ``` ### FP16 + ```bash # Accuracy bash scripts/infer_retinaface_fp16_accuracy.sh @@ -44,10 +58,10 @@ bash scripts/infer_retinaface_fp16_performance.sh ## Results -| Model | BatchSize | Precision | FPS | Easy AP(%) | Medium AP (%) | Hard AP(%) | -| :--------------: | :-------: | :-------: | :-----: | :--------: | :-----------: | :--------: | -| RetinaFace | 32 | FP16 | 5575.56 | 80.13 | 68.52 | 36.59 | +| Model | BatchSize | Precision | FPS | Easy AP(%) | Medium AP (%) | Hard AP(%) | +| :--------: | :-------: | :-------: | :------: | :--------: | :-----------: | :--------: | +| RetinaFace | 32 | FP16 | 8304.626 | 80.13 | 68.52 | 36.59 | ## Reference -Face-Detector-1MB-with-landmark: \ No newline at end of file +Face-Detector-1MB-with-landmark: diff --git a/models/cv/detection/retinaface/igie/widerface_evaluate/box_overlaps.pyx b/models/cv/detection/retinaface/igie/widerface_evaluate/box_overlaps.pyx index 66544c1ffe2e9f6919036fe2c689fe3536f89866..3cedef8f6062b53000ba52c28eb3cb8e868d9c59 100755 --- a/models/cv/detection/retinaface/igie/widerface_evaluate/box_overlaps.pyx +++ b/models/cv/detection/retinaface/igie/widerface_evaluate/box_overlaps.pyx @@ -1,6 +1,8 @@ # -------------------------------------------------------- # Fast R-CNN # Copyright (c) 2015 Microsoft +# Copyright (c) 2024, Shanghai Iluvatar CoreX Semiconductor Co., Ltd. +# All Rights Reserved. # Licensed under The MIT License [see LICENSE for details] # Written by Sergey Karayev # -------------------------------------------------------- diff --git a/models/cv/detection/rtmdet/igie/README.md b/models/cv/detection/rtmdet/igie/README.md index 737d203c279282ccbff080934e52926d7e3d301f..25b1f9b3af2be339fd701f2654641e75afe837cf 100644 --- a/models/cv/detection/rtmdet/igie/README.md +++ b/models/cv/detection/rtmdet/igie/README.md @@ -25,12 +25,17 @@ pip3 install mmengine==0.10.4 ``` ### Download -Pretrained model: https://download.openmmlab.com/mmpose/v1/projects/rtmpose/rtmdet_nano_8xb32-100e_coco-obj365-person-05d8511e.pth + +Pretrained model: Dataset: to download the validation dataset. +```bash +wget https://download.openmmlab.com/mmpose/v1/projects/rtmpose/rtmdet_nano_8xb32-100e_coco-obj365-person-05d8511e.pth +``` ### Model Conversion + ```bash # export onnx model python3 export.py --weight rtmdet_nano_8xb32-100e_coco-obj365-person-05d8511e.pth --cfg rtmdet_nano_320-8xb32_coco-person.py --output rtmdet.onnx @@ -56,6 +61,6 @@ bash scripts/infer_rtmdet_fp16_performance.sh ## Results -| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | -| :-------: | :-------: | :---------: | :-------: | :-------: | :--------: | -| RTMDet | 32 | 320x320 | FP16 | 2627.15 | 0.619 | \ No newline at end of file +| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | +| :----: | :-------: | :---------: | :-------: | :------: | :--------: | +| RTMDet | 32 | 320x320 | FP16 | 4006.849 | 0.619 | diff --git a/models/cv/pose_estimation/rtmpose/igie/README.md b/models/cv/pose_estimation/rtmpose/igie/README.md index 2b7fb929b6fdb9c0b83f883858c6efba4cf408df..02cbc5fca1e5000c9317e60b51839d421d134e6f 100644 --- a/models/cv/pose_estimation/rtmpose/igie/README.md +++ b/models/cv/pose_estimation/rtmpose/igie/README.md @@ -25,12 +25,17 @@ pip3 install mmengine==0.10.4 ``` ### Download -Pretrained model: https://download.openmmlab.com/mmpose/v1/projects/rtmposev1/rtmpose-m_simcc-aic-coco_pt-aic-coco_420e-256x192-63eb25f7_20230126.pth + +Pretrained model: Dataset: to download the validation dataset. +```bash +wget https://download.openmmlab.com/mmpose/v1/projects/rtmposev1/rtmpose-m_simcc-aic-coco_pt-aic-coco_420e-256x192-63eb25f7_20230126.pth +``` ### Model Conversion + ```bash # export onnx model python3 export.py --weight rtmpose-m_simcc-aic-coco_pt-aic-coco_420e-256x192-63eb25f7_20230126.pth --cfg rtmpose-m_8xb256-420e_coco-256x192.py --output rtmpose.onnx @@ -56,6 +61,6 @@ bash scripts/infer_rtmpose_fp16_performance.sh ## Results -| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | -| :-------: | :-------: | :---------: | :-------: | :-------: | :--------: | -| RTMPose | 32 | 252x196 | FP16 | 2313.33 | 0.936 | +| Model | BatchSize | Input Shape | Precision | FPS | mAP@0.5(%) | +| :-----: | :-------: | :---------: | :-------: | :------: | :--------: | +| RTMPose | 32 | 256x192 | FP16 | 3459.341 | 0.936 |