PoseFormerV2
PoseFormerV2 copied to clipboard
QitaoZhao
The project is an official implementation of our paper "PoseFormerV2: Exploring Frequency Domain for Efficient and Robust 3D Human Pose Estimation".
Thanks for your awesome work. I also hope to use the MPI-INF-3DHP data set in my experiment. How did you deal with MPI-INF-3DHP? This problem has been bothering me for...
https://github.com/QitaoZhao/PoseFormerV2/blob/0c0b1252f6ce722ed621d82b551362f25a023285/common/model_poseformer.py#L224 https://github.com/QitaoZhao/PoseFormerV2/blob/0c0b1252f6ce722ed621d82b551362f25a023285/common/model_poseformer.py#L120 https://github.com/QitaoZhao/PoseFormerV2/blob/0c0b1252f6ce722ed621d82b551362f25a023285/common/model_poseformer.py#L122 https://github.com/QitaoZhao/PoseFormerV2/blob/0c0b1252f6ce722ed621d82b551362f25a023285/common/model_poseformer.py#L127 https://github.com/QitaoZhao/PoseFormerV2/blob/0c0b1252f6ce722ed621d82b551362f25a023285/common/model_poseformer.py#L128 `x[:, :f//2]` `x[:, f//2:]` Are the frame numbers of x set incorrectly, or did I misunderstand? x combine with Spatial_feature , does x belongs to...
First, thanks for your greatest work of PoseFormerV2. I checked your code and found that when evaluating the mpii dataset, the calculation of the two metrics pck and auc is...
Traceback (most recent call last): File "/content/PoseFormerV2/demo/vis.py", line 286, in get_pose2D(video_path, output_dir) File "/content/PoseFormerV2/demo/vis.py", line 90, in get_pose2D keypoints, scores = hrnet_pose(video_path, det_dim=416, num_peroson=1, gen_output=True) File "/content/PoseFormerV2/demo/lib/hrnet/gen_kpts.py", line 95, in...
In requirements.txt, there is a package named install==1.3.5, its name has changed to pip-install. It should be pip-install=1.3.5 now.
Is there a test code for the MPI-INF-3DHP training model? The value is modified according to the draw_3d_keypoint_3dhp.py script Training human3.6 inference code doesn't work.
