TengdaHan / DPC

Video Representation Learning by Dense Predictive Coding. Tengda Han, Weidi Xie, Andrew Zisserman.

TengdaHan/DPC Issues

pretrained weight files aren't able to decompress
Closed 2 years ago2
Allow compatibility with latest torch (torch 18.1)
Closed 3 years ago2
Link to pretrained model on UCF101 not working
Closed 3 years ago2
Training Time
Closed 3 years ago2
The necessary of train/validation split
Closed 3 years ago1
Tracking running statistics for BN
Closed 3 years ago1
Pre-training on UCF101
Closed 4 years ago1
gray scale performance
Closed 4 years ago1
Sanity check: Classification after Self-Supervised training
Closed 4 years ago3
Batch size per GPU
Closed 4 years ago3
Experimental details of Table 1?
Closed 4 years ago6
Context and hidden state
Closed 4 years ago3
Optimizing dataloading
Closed 4 years ago2
Not understanding why you take the last sequences and not the last samples of the sequences
Closed 4 years ago4
negative examples
Closed 4 years ago1
About the Contrastive Loss
Closed 4 years ago5
Pretrained Networks missing some layer values
Closed 5 years ago2
How to compute the self-supervised acc?
Closed 5 years ago1
Possible reasons for loss function not going down
Closed 5 years ago6
Why are you shuffling the sequence of images?
Closed 5 years ago2
Using only one image at every time step instead of more than one?
Closed 5 years ago7
The accuracy of SOTA on UCF101 is more than 98%, why DPC is worse?
Closed 5 years ago3
What do you mean by sequential prediction in your paper?
Closed 5 years ago1
Code is very slow on Google Colab
Closed 5 years ago1
Dimensions [B, N, C, SL, W, H] of input block?
Closed 5 years ago2
Reproducing the paper's results on HMDB51/UCF101
Closed 5 years ago2