GAP
GAP copied to clipboard
MartinXM
official implementation for Language Supervised Training for Skeleton-based Action Recognition
Thanks for the great work! I'm working on a project that could greatly benefit from the use of your pre-trained weights as described in your paper. Would it be possible...
I want to switch the backbone to ST-GCN, but it seems that this approach is not effective. I suspect it might be because I haven't trained the CLIP. Can you...
I have run your code according your settings, but I have got the accuracy for NW-UCLA 96.12%.
Hello, when i tried to use CTR-GCN to train NW-UCLA, i got 95.5 accurancy , 1% lower than the result CTR-GCN given. Did you meet the same problem in your...
Hi, I'm wondering that if i modify codes for 2d skeleton data then GAP can work well.
您好，想请问下 https://github.com/MartinXM/CDAD 的数据集下载失效了，请问可以从哪里下载呢？ 谢谢。
Thanks for your patient and timely answer. I have another question.In Model_4part_1st,part partition strategy like this: head_list = torch.Tensor([2,3,20]).long() hand_list = torch.Tensor([4,5,6,7,8,9,10,11,21,22,23,24]).long() foot_list = torch.Tensor([12,13,14,15,16,17,18,19]).long() hip_list = torch.Tensor([0,1,2,12,16]).long() But in...
In 1st_bone.yaml, train/test_feeder_args:bone is False; and in 1st_bone_vel.yaml is the same setting but change the vel state is True.I am very confused about this,because CTR-GCN set the bone is True...