TengdaHan DPC issues - Githubissues

TengdaHan / DPC

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

MIT License

251 stars 34 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

pretrained weight files aren't able to decompress

#27 andreiamatuni closed 2 years ago
2
Minor edits to enable compatibility with recent torch versions

#26 colleenjg closed 3 years ago
1
Allow compatibility with latest torch (torch 18.1)

#25 colleenjg closed 3 years ago
2
Link to pretrained model on UCF101 not working

#24 sergiuoprea closed 3 years ago
2
Training Time

#23 praveenr2998 closed 3 years ago
2
The necessary of train/validation split

#22 Dylan-H-Wang closed 3 years ago
1
Tracking running statistics for BN

#21 YuanLiuuuuuu closed 3 years ago
1
Pre-training on UCF101

#20 Yuliang-Zou closed 4 years ago
1
gray scale performance

#19 YuqiHUO closed 4 years ago
1
Sanity check: Classification after Self-Supervised training

#18 DAVEISHAN closed 4 years ago
3
Batch size per GPU

#17 ruoshiliu closed 4 years ago
3
Context and hidden state

#16 ruoshiliu closed 4 years ago
3
Optimizing dataloading

#15 ruoshiliu closed 4 years ago
2
Not understanding why you take the last sequences and not the last samples of the sequences

#14 wolhandlerdeb closed 4 years ago
4
negative examples

#13 wolhandlerdeb closed 4 years ago
1
About the Contrastive Loss

#12 lixiangyu-1008 closed 4 years ago
5
Experimental details of Table 1?

#11 jiujing23333 closed 4 years ago
6
Pretrained Networks missing some layer values

#10 cfunk1210 closed 4 years ago
2
How to compute the self-supervised acc?

#9 gaobb closed 4 years ago
1
Why are you shuffling the sequence of images?

#8 sarimmehdi closed 4 years ago
2
Possible reasons for loss function not going down

#7 sarimmehdi closed 4 years ago
6
The accuracy of SOTA on UCF101 is more than 98%, why DPC is worse?

#6 PGogo closed 5 years ago
3
Using only one image at every time step instead of more than one?

#5 sarimmehdi closed 5 years ago
7
What do you mean by sequential prediction in your paper?

#4 sarimmehdi closed 5 years ago
1
Code is very slow on Google Colab

#3 sarimmehdi closed 5 years ago
1
Dimensions [B, N, C, SL, W, H] of input block?

#2 sarimmehdi closed 5 years ago
2
Reproducing the paper's results on HMDB51/UCF101

#1 mengyuest closed 5 years ago
2