Giters
kenshohara
/
3D-ResNets-PyTorch
3D ResNets for Action Recognition (CVPR 2018)
Geek Repo:
Geek Repo
Github PK Tool:
Github PK Tool
Stargazers:
3835
Watchers:
59
Issues:
269
Forks:
929
kenshohara/3D-ResNets-PyTorch Issues
datasets used in model training
Updated
2 months ago
Comments count
1
AssertionError when I inference
Updated
2 months ago
Comments count
1
accuracy = sum(correct)/n_ground_truth : ZeroDivision Error: Division by Zero
Updated
2 months ago
Comments count
2
ABOUT ucf101_json
Updated
3 months ago
Comments count
1
how to change the number of clips generated arround the selected temporal position ?
Updated
4 months ago
How to calculate UCF101's top1 video level accuracy
Closed
2 years ago
transforms.Scale renamed transforms.Resize
Updated
4 months ago
Comments count
1
How big CPU/GPU memory need to run 3D ResNext-101
Updated
7 months ago
Comments count
4
error : expected tensor as element 0 in argument 0
Updated
9 months ago
evaluate top-1 video accuracy of recognition result : invalid argument for the ground truth ?
Updated
9 months ago
KeyError: 'Biking' when training on UCF101
Closed
3 years ago
Comments count
2
what does the code line: "r3d200_KM_200ep.pth: --model resnet --model_depth 200 --n_pretrain_classes 1039" mean ?
Updated
a year ago
Comments count
2
RuntimeError: view size is not compatible with input tensor's size and stride (at least one dimension spans across two contiguous subspaces). Use .reshape(...) instead.
Updated
a year ago
Comments count
2
main.py: error: unrecognized arguments: hmdb51_3.json
Updated
a year ago
Comments count
2
is there 3d densenet pretrained models?
Updated
2 years ago
Image Resolution 112*112
Updated
2 years ago
Comments count
1
train on ucf,"ValueError: num_samples should be a positive integer value, but got num_samples=0"
Updated
2 years ago
Comments count
3
Why is opt.n_val_samples 3
Updated
2 years ago
unable to create val.json
Updated
2 years ago
Comments count
4
Running on Windows, convert avi to jpg. FileNotFoundError: [WinError 2] The system cannot find the specified file.
Closed
2 years ago
Comments count
2
Running on Windows, convert avi to jpg. FileNotFoundError: [WinError 2] The system cannot find the specified file.
Updated
2 years ago
Comments count
4
the solution about the low acc in UCF101_01 after loading pretrained-model
Closed
2 years ago
How to convert a model(save_200.pth) to an onnx model?
Updated
3 years ago
UCF-101 - Convert from avi to jpg - estimated run time
Updated
3 years ago
Comments count
4
pretrained model on IG-65M or IG-65M+Kinetics
Updated
3 years ago
Accimage is not available
Closed
3 years ago
Comments count
1
Accuracy for ntu60 RGB
Updated
3 years ago
ResNeXt101-kinetics on HMDB51-split02,split-03; and UCF101-split02,split03 pretrained model
Updated
3 years ago
questions about old pretrained models
Updated
3 years ago
Comments count
1
Error when load state_dict to finetune on ucf101 from pretrained model r2p1d34_K_200ep.pth : Missing key(s) in state_dict
Updated
3 years ago
Comments count
2
training from scratch
Closed
3 years ago
Question About Batch_size and Epoch
Updated
3 years ago
input_shape
Updated
3 years ago
module.***
Updated
3 years ago
Data Augmentation Question
Updated
3 years ago
Bug in TemporalEvenCrop
Updated
3 years ago
Slow dataloading every N batches, where N=num_threads
Updated
3 years ago
Comments count
2
use model test single video
Updated
3 years ago
Comments count
1
Test model
Updated
3 years ago
Is there a simple way to get inference results on video files that are not part of the considered data sets?
Updated
3 years ago
Comments count
2
visualization!!!
Updated
3 years ago
Comments count
1
Training for new videos.
Updated
3 years ago
generate_video_jpgs.py arguments are required error
Closed
3 years ago
Comments count
2
KeyError of distributed training
Closed
3 years ago
Comments count
2
GPU utilization rate and distribution training
Updated
3 years ago
Comments count
1
error in generating ucf101.json
Updated
3 years ago
Comments count
1
Very poor performance on pre-trained models.. maybe un matching Labels to index file
Closed
3 years ago
Comments count
1
Train time on UFC101 fine tuning on fully connected layer and loading model query
Closed
3 years ago
Comments count
1
avi failed to generate jpg
Closed
3 years ago
Comments count
1
I inflated the google's lasted lightweighted 2d cnn model: MobileNetv3 to 3d, and get a higher precision 52% on ucf101 from scratch
Closed
3 years ago
Previous
Next