Merge pull request open-mmlab#1 from open-mmlab/wc/modelzoo

wjkim81 · Jul 10, 2020 · 4336e90 · 4336e90
2 parents ee44566 + 9027dc3
commit 4336e90
Show file tree

Hide file tree

Showing 5 changed files with 39 additions and 18 deletions.
diff --git a/configs/top_down/hrnet/coco/hrnet_w32_coco_256x192.py b/configs/top_down/hrnet/coco/hrnet_w32_coco_256x192.py
@@ -99,7 +99,7 @@
     bbox_thr=1.0,
     use_gt_bbox=True,
     image_thr=0.0,
-    bbox_file='pretrained_models/det/'
+    bbox_file='data/coco/person_detection_results/'
     'COCO_val2017_detections_AP_H_56_person.json',
 )
 

diff --git a/configs/top_down/resnet/coco/res50_coco_256x192.py b/configs/top_down/resnet/coco/res50_coco_256x192.py
@@ -69,7 +69,7 @@
     bbox_thr=1.0,
     use_gt_bbox=True,
     image_thr=0.0,
-    bbox_file='pretrained_models/det/'
+    bbox_file='data/coco/person_detection_results/'
     'COCO_val2017_detections_AP_H_56_person.json',
 )
 

diff --git a/docs/benchmark.md b/docs/benchmark.md
@@ -22,14 +22,15 @@ The mAP (the mean average precision) is used as the evaluation metrics.
 | resnet_50  | 256x192  | **1.10** | 1.12 | **0.718** | 0.704 |
 | resnet_50  | 384x288  | **1.83** | 2.15 | **0.731** | 0.722 |
 | resnet_101 | 256x192  | **1.10** | 1.46 | **0.726** | 0.714 |
-| resnet_101 | 384x288  | **1.89** | 3.08 | **xxxxx** | 0.736 |
-| resnet_152 | 256x192  | **1.11** | 1.71 | **xxxxx** | 0.720 |
+| resnet_101 | 384x288  | **1.89** | 3.08 | **0.748** | 0.736 |
+| resnet_152 | 256x192  | **1.11** | 1.71 | **0.735** | 0.720 |
 | resnet_152 | 384x288  | **1.85** | 3.71 | **0.750** | 0.743 |
 | hrnet_w32  | 256x192  | **1.11** | 2.09 | **0.746** | 0.744 |
 | hrnet_w32  | 384x288  | **1.70** | 3.36 | **0.760** | 0.758 |
 | hrnet_w48  | 256x192  | **1.08** | 2.28 | **0.756** | 0.751 |
 | hrnet_w48  | 384x288  | **1.84** | 3.46 | **0.767** | 0.763 |
 
+
 ## Hardware
 
 - 8 NVIDIA Tesla V100 (32G) GPUs

diff --git a/docs/getting_started.md b/docs/getting_started.md
@@ -41,7 +41,7 @@ mmpose
 For using custom datasets, please refer to [Tutorial 2: Adding New Dataset](tutorials/new_dataset.md)
 
 ## Prepare Pratrained Models
-Download imagenet pretrained models from our [model zoo](www..)
+Download imagenet pretrained models from our [model zoo](model_zoo.md)
 
 ```
 mmpose

diff --git a/docs/model_zoo.md b/docs/model_zoo.md
@@ -1,17 +1,37 @@
 # Model Zoo
+
 ## Result
 
-### Top-Down
+### Results on COCO val2017 with detector having human AP of 56.4 on COCO val2017 dataset
+
+| Arch | Input Size | AP | AP<sup>50</sup> | AP<sup>75</sup> | AR | AR<sup>50</sup> | ckpt | log |
+| :----------------- | :-----------: | :------: | :------: | :------: | :------: | :------: |:------: |:------: |
+| pose_resnet_50  | 256x192 | 0.718 | 0.898 | 0.795 | 0.773 | 0.937 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res50_coco_256x192-ec54d7f3_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res50_coco_256x192_20200709.log.json) |
+| pose_resnet_50  | 384x288 | 0.731 | 0.900 | 0.799 | 0.783 | 0.931 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res50_coco_384x288-e6f795e9_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res50_coco_384x288_20200709.log.json) |
+| pose_resnet_101 | 256x192 | 0.726 | 0.899 | 0.806 | 0.781 | 0.939 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res101_coco_256x192-6e6babf0_20200708.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res101_coco_256x192_20200708.log.json) |
+| pose_resnet_101 | 384x288 | 0.748 | 0.905 | 0.817 | 0.798 | 0.940 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res101_coco_384x288-8c71bdc9_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res101_coco_384x288_20200709.log.json) |
+| pose_resnet_152 | 256x192 | 0.735 | 0.905 | 0.812 | 0.790 | 0.943 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res152_coco_256x192-f6e307c2_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res152_coco_256x192_20200709.log.json) |
+| pose_resnet_152 | 384x288 | 0.750 | 0.908 | 0.821 | 0.800 | 0.942 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res152_coco_384x288-3860d4c9_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res152_coco_384x288_20200709.log.json) |
+| pose_hrnet_w32  | 256x192 | 0.746 | 0.904 | 0.819 | 0.799 | 0.942 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w32_coco_256x192-c78dce93_20200708.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w32_coco_256x192_20200708.log.json) |
+| pose_hrnet_w32  | 384x288 | 0.760 | 0.906 | 0.829 | 0.810 | 0.943 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w32_coco_384x288-d9f0d786_20200708.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w32_coco_384x288_20200708.log.json) |
+| pose_hrnet_w48  | 256x192 | 0.756 | 0.907 | 0.825 | 0.806 | 0.942 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w48_coco_256x192-b9e0b3ab_20200708.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w48_coco_256x192_20200708.log.json) |
+| pose_hrnet_w48  | 384x288 | 0.767 | 0.910 | 0.831 | 0.816 | 0.946 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w48_coco_384x288-314c8528_20200708.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hrnet/hrnet_w48_coco_384x288_20200708.log.json) |
+| pose_scnet_50   | 256x192 | 0.728 | 0.899 | 0.807 | 0.784 | 0.938 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet50_coco_256x192-6920f829_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet50_coco_256x192_20200709.log.json) |
+| pose_scnet_50  | 384x288 | 0.751 | 0.906 | 0.818 | 0.802 | 0.943 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet50_coco_384x288-9cacd0ea_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet50_coco_384x288_20200709.log.json) |
+| pose_scnet_101  | 256x192 | 0.733 | 0.903 | 0.813 | 0.790 | 0.941 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet101_coco_256x192-6d348ef9_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet101_coco_256x192_20200709.log.json) |
+| pose_scnet_101  | 384x288 | 0.752 | 0.906 | 0.823 | 0.804 | 0.943 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet101_coco_384x288-0b6e631b_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/scnet/scnet101_coco_384x288_20200709.log.json) |
+| dark_pose_resnet_50 | 256x192 | 0.724 | 0.898 | 0.800 | 0.777 | 0.936 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res50_coco_256x192_dark-43379d20_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/resnet/res50_coco_256x192_dark_20200709.log.json) |
+| pose_hourglass_52 | 256x256 | 0.726 | 0.896 | 0.799 | 0.780 | 0.934 | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hourglass/hourglass52_coco_256x256-4ec713ba_20200709.pth) | [log](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/top_down/hourglass/hourglass52_coco_256x256_20200709.log.json) |
+
+
+### Pretrained backbones on ImageNet
 
-| Arch | Input Size | AP | AP<sup>50</sup> | AP<sup>75</sup> | AR | AR<sup>50</sup> |
-| :----------------- | :-----------: | :------: | :------: | :------: | :------: | :------: |
-| pose_resnet_50 | 256x192 | 0.734 | 0.925 | 0.814 | 0.764 | 0.936 |
-| pose_resnet_50 | 384x288 | 0.746 | 0.925 | 0.811 | 0.777 | 0.936 |
-| pose_resnet_101 | 256x192 | 0.745 | 0.925 | 0.825 | 0.776 | 0.940 |
-| pose_resnet_101 | 384x288 | 0.770 | 0.934 | 0.845 | 0.798 | 0.942 |
-| pose_resnet_152 | 256x192 | 0.755 | 0.935 | 0.835 | 0.784 | 0.942 |
-| pose_resnet_152 | 384x288 | 0.773 | 0.936 | 0.848 | 0.803 | 0.949 |
-| pose_hrnet_w32 | 256x192 | 0.767 | 0.936 | 0.737 | 0.794 | 0.945 |
-| pose_hrnet_w32 | 384x288 | 0.783 | 0.936 | 0.848 | 0.808 | 0.948 |
-| pose_hrnet_w48 | 256x192 | 0.775 | 0.936 | 0.847 | 0.803 | 0.947 |
-| pose_hrnet_w48 | 384x288 | 0.784 | 0.936 | 0.849 | 0.811 | 0.945 |
+| Arch |  ckpt |
+| :----------------- | :-----------: |
+| resnet50-19c8e357.pth | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/resnet50-19c8e357.pth)
+| resnet101-5d3b4d8f.pth | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/resnet101-5d3b4d8f.pth)
+| resnet152-b121ed2d.pth | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/resnet152-b121ed2d.pth)
+| hrnet_w32-36af842e.pth  | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/hrnet_w32-36af842e.pth) |
+| hrnet_w48-8ef0771d.pth | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/hrnet_w48-8ef0771d.pth)
+| scnet50-7ef0a199.pth | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/scnet50-7ef0a199.pth)
+| scnet101-94250a77.pth | [ckpt](https://openmmlab.oss-cn-hangzhou.aliyuncs.com/mmpose/pretrain_models/scnet101-94250a77.pth)