VIBE
VIBE copied to clipboard
mkocabas
Official implementation of CVPR2020 paper "VIBE: Video Inference for Human Body Pose and Shape Estimation"
@mkocabas Thanks for this wonderful effort. I am able to get the FBX output for my videos. However, there is one key issue. The character is always centred at the...
i run the scripts 
Does the function ``` def extract_features(model, video, bbox, debug=False, batch_size=200, kp_2d=None, dataset=None, scale=1.3): ``` in ``` lib/data_utils/feature_extractor.py ``` Can handle different image sizes? For example, can we use the image...
Hello, I have a question about the evaluation of human3.6m dataset. I used your pertained model to evaluate on the dataset, the results of 3DPW and MPI-INF-3DHP match the result...
 In your paper：Specifically, for each frame we predict the body shape parameters. Then, we apply average pooling to get a single shape (βˆ) across the whole input sequence。 But...
Hello, I have a basic question I was hoping you could help me with. from the article: _Here, past information can help resolve and constrain the pose estimate. The temporal...
Can someone please help me understand where can I get this 3dpw_test_db.pt file? Thank you
Hi, I am having trouble using the bounding boxes for the 3DPW dataset. The crop and the overlayed 2d keypoints aren't aligned. Here is a video showing what I am...
First, kp_2d have been converted to 'common': https://github.com/mkocabas/VIBE/blob/master/lib/data_utils/penn_action_utils.py#L80 Second, it still is used with 'pennaction'? https://github.com/mkocabas/VIBE/blob/master/lib/dataset/dataset_2d.py#L70
I find a error in TemporalSMPLify codes. line 124 in lib/utils/demo_utils.py : `gt_keypoints_2d_orig = j2d` First, running compute reprojection error of the network in the function smplify.get_fitting_loss. line 235 in...
