.npy - feature extract file
DungVo1507 opened this issue · 6 comments
Thank you so much for creating this useful repo!
After I follow the instructions in the Usage section but the output I get in /temp/ is the frames of the video.
Please tell me how to get a .npy file and shape (10, n/16, 2048) like in this paper Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learning.
Looking forward to your help
同问
Thank you so much for creating this useful repo!
After I follow the instructions in the Usage section but the output I get in /temp/ is the frames of the video.
Please tell me how to get a .npy file and shape (10, n/16, 2048) like in this paper Weakly-supervised Video Anomaly Detection with Robust Temporal Feature Magnitude Learning.Looking forward to your help
The frames in the temp folder is not the output, the .npy file is generated along with the temp folder with 1 * n/16 * 2048 dimension.
You can see it here
I3D_Feature_Extraction_resnet/main.py
Line 36 in 340bbec
Thank you so much for helping me,
Below is the output after I run the command line: !python main.py --datasetpath=samplevideos/ --outputpath=output:
Received Pretrained model..
Generating for samplevideos/Abuse001_x264.mp4
Preprocessing done..
batchsize 32
^C
It doesn't print "Obtained features of size:" and "done in {0}." like lines 37 and 39 in main.py. This means I still haven't created the .npy.
Can you explain what problem I'm having?
Thank you!
It seems that it is exiting inside the run function of extract_features.py. I would suggest you to keep debug points(or print statements) in extract_features.py and check where it is exiting
After running !python main.py --datasetpath=samplevideos/ --outputpath=output
The code is exiting after showing
Received Pretrained model..
Can you help me with this regard?
After executing command python main.py --datasetpath=samplevideos/ --outputpath=output
The code is exiting after showing
Received Pretrained model..
Can you help me with this regard?