kenziyuliu MS-G3D issues

kenziyuliu / MS-G3D

[CVPR 2020 Oral] PyTorch implementation of "Disentangling and Unifying Graph Convolutions for Skeleton-Based Action Recognition"

https://arxiv.org/abs/2003.14111

MIT License

430 stars 96 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Update README.md

#59 yunxi-wd closed 2 months ago
0
Shape error while predicting on custom video

#58 Zepharchit opened 1 year ago
0
Error in generating data

#57 whisper0411 opened 1 year ago
1
How to make the frames not change?

#56 LIUQI-creat closed 1 year ago
2
Multiprocess NTU-RGB+D 60 data generation

#55 alirezadizaji closed 1 year ago
1
Half Precision Training may cause discrepancies of test results between training and test time for the same model

#54 zhixuany closed 2 years ago
2
About over-smoothing

#53 15762260991 closed 2 years ago
8
ValueError: invalid literal for int() with base 10: 'ntu' when run python ntu_gendata.py

#52 CyWang-Jade closed 2 years ago
2
Change number of people in model

#51 tucachmo2202 closed 2 years ago
2
Fixed typo in global_step number acquisition

#50 PINTO0309 closed 2 years ago
1
How to use it to predict an RGB video？

#49 hellogit-0 closed 2 years ago
1
kinect 400 dataset

#48 yanghhx closed 2 years ago
1
running_mean should contain 54 elements not 108

#47 llgzr closed 2 years ago
2
Can you provide Flops of this model?

#46 wkxurn closed 3 years ago
2
Vs code运行代码直接终止，并且没有报错

#45 MosasaurYi closed 2 years ago
7
使用VS code运行直接终止并且没有报错

#44 MosasaurYi closed 3 years ago
0
ST-private

#43 satorutada closed 3 years ago
3
Softmax scores

#42 KoubaPetr closed 3 years ago
5
Recognizing activities using your library

#41 fspegni closed 3 years ago
2
Function 'CudnnBatchNormBackward' returned nan values in its 0th output

#40 saniazahan closed 3 years ago
3
数据类型为NoneTyoe格式

#39 1330805824 closed 3 years ago
2
CUDA out of memory while evaluating pretrained

#38 fspegni closed 3 years ago
4
Is the format of data generated with `ntu_gendata.py` same as described in `lshiwjx/2s-AGCN`

#37 m-e-l-u-h-a-n closed 3 years ago
0
Function 'LogSoftmaxBackward' returned nan values in its 0th output

#36 saniazahan closed 3 years ago
4
support distributed manner

#35 CodesFarmer closed 3 years ago
1
Snippet for confusion matrix

#34 egrassl closed 3 years ago
1
Some problems with motion recognition

#33 melon-coder closed 3 years ago
1
"kinetics_gendata.py" Error

#32 yaki-toki closed 3 years ago
1
required time for training on kinetics

#31 parvanehabi closed 3 years ago
1
code for plotting Fig 2 in the paper

#30 a14en9 closed 3 years ago
1
Problem in splitting actions classes

#29 najamsaleem closed 3 years ago
2
CUDA out of memory

#28 guerrifrancesco closed 3 years ago
15
Failing to reproduce results on Kinetics 400

#27 andreashaukeland closed 3 years ago
1
Result on Kinetics Skeleton 400

#26 Viozer closed 4 years ago
5
Cannot reproduce the pretrained model accuracy for NTU120 无法复现提供的pretrained模型的accuracy

#25 kfzyqin closed 4 years ago
1
what is detailed operation of collapse window reshape and fc

#24 nianniana closed 4 years ago
1
Regarding the calculation of multi-scale graphics when disentangling is set to False.

#23 a14en9 closed 4 years ago
4
Transfer Learning

#22 egrassl closed 4 years ago
1
IndexError on ensemble.py

#21 egrassl closed 4 years ago
8
Error when trying to reproduce result from pretrained model -- MKL_THREADING_LAYER

#20 saniazahan closed 4 years ago
1
Snapshots of actions

#19 hulianyuyy closed 4 years ago
2
Activity recognition using your model

#18 ywang0218 closed 4 years ago
1
How many epochs do we need to train in order to get the accuracy of the given pretrained model?

#17 kaiser-hamid-rabbi closed 4 years ago
3
About the flops of the model

#16 xf1994 closed 4 years ago
4
Some questions about table 2

#15 liubl1217 closed 4 years ago
8
Cannot reproduce the result on NTU120

#14 Viozer closed 4 years ago
12
Why not apply residual in MS-G3D or MS-GCN?

#13 tracer9 closed 4 years ago
1
Can not reproduced the results reported in the paper

#12 qiexing closed 4 years ago
1
The accuracy is so high, so will this paper be the end of NTU-60? Or will there be any potential for promotion?

#11 desertfex closed 4 years ago
11
Is this a real-time method of action recognition?

#10 JoshaPan closed 4 years ago
3