yjxiong / tsn-pytorch

Temporal Segment Networks (TSN) in PyTorch

yjxiong/tsn-pytorch Issues

报错求助magic_number = pickle_module.load(f, **pickle_load_args) _pickle.UnpicklingError: invalid load key, '<'.（困扰我很久了救救孩子吧））
Updated 5 months ago2
TypeError: forward() missing 1 required positional argument: 'input_tensor'
Updated 9 months ago3
mean() received an invalid combination of arguments - got (Tensor, list, keepdim=bool), but expected one of: * (Tensor input) * (Tensor input, torch.dtype dtype) * (Tensor input, int dim, torch.dtype dtype, Tensor out) * (Tensor input, int dim, bool keepdim, torch.dtype dtype, Tensor out) * (Tensor input, int dim, bool keepdim, Tensor out)
Updated 10 months ago
Can you provide us the pytorch code for topK consensus and attention weight pooling consensus?
Updated a year ago2
f**k-est implementation
Updated 2 years ago3
what is the command of RGB and Flow modality fusion? Thanks！！
Updated 2 years ago3
请问经过模型后的输出是什么格式呢？
Updated 2 years ago2
size mismatch for weights and bias
Updated 2 years ago4
size mistmatch?
Updated 2 years ago3
Need pre-trained ActivityNet weights for Pytorch
Updated 3 years ago
Why subtract 'new_length' to calculate 'average_duration' ?
Updated 3 years ago
training tsn on kinetic400 is ok, but validating is very low. do you test your model on the kinetics400 data set
Closed 3 years ago4
cannot downgrade pytorch to solve mismatch
Closed 4 years ago2
RuntimeError: Legacy autograd function
Updated 3 years ago1
How to fuse the scores from both RGB and FLOW on training and testing?
Closed 4 years ago1
Training the TSN model on custom dataset - couldn't implement as mentioned in paper
Updated 3 years ago1
what is the command of RGB and Flow modality fusion?
Closed 4 years ago2
video live test?
Updated 3 years ago
What should I do if I want to change the main structure of the network to InceptionV3?
Closed 5 years ago1
Solved
Closed 3 years ago
No module named 'mmaction.datasets.utils'
Updated 3 years ago1
Issue about training Kinetics400
Updated 3 years ago6
The score of rgb and flow
Updated 3 years ago
getting 84% in ucf101 on split 1
Updated 3 years ago2
RuntimeError: bool value of non-empty torch.cuda.ByteTensor objects is ambiguous
Updated 3 years ago
error about epoch print statement when i use the main.py to train with ucf101 dataset
Closed 3 years ago1
weried phenomenon about using main.py train ucf101 dataset
Closed 4 years ago1
Error: return int(self._data[1]) IndexError: list index out of range
Closed 4 years ago
Feature Extraction
Updated 4 years ago
Test time increase linearly then crashed for RecursionError
Updated 4 years ago
would you share optical flow and frame for the paper
Updated 4 years ago
model not find
Updated 4 years ago
Can I run build_file_list.py with gpu, for extract flow is too slow
Closed 4 years ago
how to create RGB <ucf101_rgb_val_list> ?
Updated 4 years ago3
AttributeError: 'BNInception' object has no attribute 'fc
Updated 4 years ago2
Where can l find the pre-trained models for UCF-101 and HMDB-51 ?
Updated 4 years ago1
problem about VGG module
Closed 4 years ago
Did anyone reproduce TSN on ActivityNet v1.3? Or are there any pretrained models on ActivityNet v1.3 available?
Updated 4 years ago
If I set the batchsize to 16, which parameters should I change accordingly, thank you very much.
Updated 5 years ago2
test_model.py IndexError: list index out of range
Closed 5 years ago1
how can i compute the speed
Updated 5 years ago
RuntimeError: Error(s) in loading state_dict for BNInception: size mismatch
Closed 5 years ago2
.npz flow score file of hmdb51 and ucf101?
Updated 5 years ago1
about flow length
Closed 5 years ago1
About data prepare
Closed 5 years ago2
where is the pretrained args "weights" defined? I can't find it.
Updated 5 years ago
How the file structure of train sets looks like?
Closed 5 years ago2
How to do modality fusion
Closed 5 years ago4
issue about fushion tools/eval_scores.py ValueError: operands could not be broadcast
Closed 5 years ago1
Issue in test_models.py
Updated 5 years ago