## 2D Human Whole-Body Pose Demo

<img src="https://user-images.githubusercontent.com/9464825/95552839-00a61080-0a40-11eb-818c-b8dad7307217.gif" width="600px" alt><br>

### 2D Human Whole-Body Pose Top-Down Image Demo

#### Using gt human bounding boxes as input

We provide a demo script to test a single image, given gt json file.

```shell
python demo/top_down_img_demo.py \
    ${MMPOSE_CONFIG_FILE} ${MMPOSE_CHECKPOINT_FILE} \
    --img-root ${IMG_ROOT} --json-file ${JSON_FILE} \
    --out-img-root ${OUTPUT_DIR} \
    [--show --device ${GPU_ID or CPU}] \
    [--kpt-thr ${KPT_SCORE_THR}]
```

Examples:

```shell
python demo/top_down_img_demo.py \
    configs/wholebody/2d_kpt_sview_rgb_img/topdown_heatmap/coco-wholebody/hrnet_w48_coco_wholebody_384x288_dark_plus.py \
    https://download.openmmlab.com/mmpose/top_down/hrnet/hrnet_w48_coco_wholebody_384x288_dark-f5726563_20200918.pth \
    --img-root tests/data/coco/ --json-file tests/data/coco/test_coco.json \
    --out-img-root vis_results
```

To run demos on CPU:

```shell
python demo/top_down_img_demo.py \
    configs/wholebody/2d_kpt_sview_rgb_img/topdown_heatmap/coco-wholebody/hrnet_w48_coco_wholebody_384x288_dark_plus.py \
    https://download.openmmlab.com/mmpose/top_down/hrnet/hrnet_w48_coco_wholebody_384x288_dark-f5726563_20200918.pth \
    --img-root tests/data/coco/ --json-file tests/data/coco/test_coco.json \
    --out-img-root vis_results \
    --device=cpu
```

#### Using mmdet for human bounding box detection

We provide a demo script to run mmdet for human detection, and mmpose for pose estimation.

Assume that you have already installed [mmdet](https://github.com/open-mmlab/mmdetection).

```shell
python demo/top_down_img_demo_with_mmdet.py \
    ${MMDET_CONFIG_FILE} ${MMDET_CHECKPOINT_FILE} \
    ${MMPOSE_CONFIG_FILE} ${MMPOSE_CHECKPOINT_FILE} \
    --img-root ${IMG_ROOT} --img ${IMG_FILE} \
    --out-img-root ${OUTPUT_DIR} \
    [--show --device ${GPU_ID or CPU}] \
    [--bbox-thr ${BBOX_SCORE_THR} --kpt-thr ${KPT_SCORE_THR}]
```

Examples:

```shell
python demo/top_down_img_demo_with_mmdet.py \
    demo/mmdetection_cfg/faster_rcnn_r50_fpn_coco.py \
    https://download.openmmlab.com/mmdetection/v2.0/faster_rcnn/faster_rcnn_r50_fpn_1x_coco/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth \
    configs/wholebody/2d_kpt_sview_rgb_img/topdown_heatmap/coco-wholebody/hrnet_w48_coco_wholebody_384x288_dark_plus.py \
    https://download.openmmlab.com/mmpose/top_down/hrnet/hrnet_w48_coco_wholebody_384x288_dark-f5726563_20200918.pth \
    --img-root tests/data/coco/ \
    --img 000000196141.jpg \
    --out-img-root vis_results
```

### 2D Human Whole-Body Pose Top-Down Video Demo

We also provide a video demo to illustrate the results.

Assume that you have already installed [mmdet](https://github.com/open-mmlab/mmdetection).

```shell
python demo/top_down_video_demo_with_mmdet.py \
    ${MMDET_CONFIG_FILE} ${MMDET_CHECKPOINT_FILE} \
    ${MMPOSE_CONFIG_FILE} ${MMPOSE_CHECKPOINT_FILE} \
    --video-path ${VIDEO_FILE} \
    --out-video-root ${OUTPUT_VIDEO_ROOT} \
    [--show --device ${GPU_ID or CPU}] \
    [--bbox-thr ${BBOX_SCORE_THR} --kpt-thr ${KPT_SCORE_THR}]
```

Examples:

```shell
python demo/top_down_video_demo_with_mmdet.py \
    demo/mmdetection_cfg/faster_rcnn_r50_fpn_coco.py \
    https://download.openmmlab.com/mmdetection/v2.0/faster_rcnn/faster_rcnn_r50_fpn_1x_coco/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth \
    configs/wholebody/2d_kpt_sview_rgb_img/topdown_heatmap/coco-wholebody/hrnet_w48_coco_wholebody_384x288_dark_plus.py \
    https://download.openmmlab.com/mmpose/top_down/hrnet/hrnet_w48_coco_wholebody_384x288_dark-f5726563_20200918.pth \
    --video-path https://user-images.githubusercontent.com/87690686/137440639-fb08603d-9a35-474e-b65f-46b5c06b68d6.mp4 \
    --out-video-root vis_results
```

### Speed Up Inference

Some tips to speed up MMPose inference:

For top-down models, try to edit the config file. For example,

1. set `flip_test=False` in [pose_hrnet_w48_dark+](https://github.com/open-mmlab/mmpose/tree/e1ec589884235bee875c89102170439a991f8450/configs/wholebody/darkpose/coco-wholebody/hrnet_w48_coco_wholebody_384x288_dark_plus.py#L80).
1. set `post_process='default'` in [pose_hrnet_w48_dark+](https://github.com/open-mmlab/mmpose/tree/e1ec589884235bee875c89102170439a991f8450/configs/wholebody/darkpose/coco-wholebody/hrnet_w48_coco_wholebody_384x288_dark_plus.py#L81).
1. use faster human bounding box detector, see [MMDetection](https://mmdetection.readthedocs.io/en/latest/model_zoo.html).