Open hundanLi opened 4 years ago
I retrain it replacing the line with
kps = normalize_screen_coordinates(kps[..., :2], w=cam['res_w'], h=cam['res_h'])
and it works well. Thank you for your work!
In the line https://github.com/facebookresearch/VideoPose3D/blob/bc7db2c378cac881d324bc2b3d4b2b404b0dc619/run.py#L178 poses_valid_2d[0].shape[-1] = 2, however the shape of the keypoints I get by detectron is [frames, 17, 3]. I has followed the author's suggestions to uncomment https://github.com/facebookresearch/VideoPose3D/blob/cf94b42e1edf144f8eff2da0f43be737225e43bd/data/data_utils.py#L81 , so I get keypoints whose shape is [frames, 17, 3]. Can I just use kps[..., :2] to get it work?
When I finish training the inference module
pretrained_h36m_detectron_coco.bin
using:and then test on a custom
.npz
predicted by detectron2 following https://github.com/darkAlert/VideoPose3d_with_Detectron2 using:it occures a size mismatch error while loading module:
How can I address this problem, please?