3D-ResNets-PyTorch

3D-ResNets-PyTorch copied to clipboard

I found that when fine-tuning UCF101, split1 partition was used to verify that the number in the dataset was 11349 instead of 3783, and why was batchsize opt.batch_size // opt.n_val_samples?

YTHmamba

unable to create val.json

4

Hi, Executed the code with the default dataset. Trained the model with kinetics dataset after successful training. Unable to find a val.json for evaluation. Please let me know how to...

suravijayjilla

About the accuracy of 94.5% on UCF101 using the Resnext101?

27

Hello, can you tell me how can achieve the the accuracy of 94.5% on UCF101 using the Resnext101? I use your code, the same network architecture(Resnext101) and your pretrained parameters(resnext-101-64f-kinetics-UCF101_split1.pth)...

Wei2Huang

Is the train.log right?

4

@kenshohara First , it is a great job! I use the kinetics data to train a resnet-34-kinetics model. Every action have 50 MP4s. I train the model just like this:...

zy486at189cn

when fine-tune the pretrained model from kinetics on ucf101 or hmdb51, the accuracy is lower than it in your paper about 7, 8 percent?

5

huangwei1995

Question about Training about ResNet-50

8

Hello, I was trying to train resnet-50 from scratch on UCF101 split 1. but my validation accuracy(clip) is only about 40%. The accuracy is also 40% on resnet-18 which is...

amm040341

Running on Windows, convert avi to jpg. FileNotFoundError: [WinError 2] The system cannot find the specified file.

4

**(py36) E:\opt\workspace\3D-ResNets-PyTorch-master\util_scripts>python -m generate_video_jpgs \opt\workspace\3D-ResNets-PyTorch-master\data\UCF-101 \opt\workspace\3D-ResNets-PyTorch-master\data\UCF-jpg ucf101** Traceback (most recent call last): File "generate_video_jpgs.py", line 117, in for class_dir_path in class_dir_paths) File "E:\opt\Anaconda3\envs\py36\lib\site-packages\joblib\parallel.py", line 1054, in __call__ self.retrieve() File "E:\opt\Anaconda3\envs\py36\lib\site-packages\joblib\parallel.py",...

N00k0428

Why using activitynet's mean to normalize?

5

question 1: we need to train kinetices model first, however why using activitynet dataset mean to normlize? in opts.py the "mean_dataset' is 'activitynet' by default, In the next the opt.mean...

tanxjtu

RuntimeError: cuda runtime error (59) : device-side assert triggered at /pytorch/torch/lib/THC/generated/../THCReduceAll.cuh:339 terminate called after throwing an instance of 'std::runtime_error' what(): cuda runtime error (59) : device-side assert triggered at /pytorch/torch/lib/THC/generic/THCStorage.c:184

15

dataset loading [0/3570] dataset loading [1000/3570] dataset loading [2000/3570] dataset loading [3000/3570] dataset loading [0/1530] dataset loading [1000/1530] run train at epoch 1 Epoch: [1][1/112] Time 4.807 (4.807) Data 2.836...

vateye

AssertionError when I inference

1

I used the r2p1d18_K_200ep.pth and finetune it on hmdb51 dataset,and when I want to use it to inference there is an AssertionError `CUDA_VISIBLE_DEVICES=0,1,2,3 python main.py --root_path /home/pubNAS/jianfei/3D-ResNets-PyTorch-master/data --video_path hmdb51-videos/jpg --annotation_path...

z369437558