wtalc-pytorch
wtalc-pytorch copied to clipboard
sujoyp
W-TALC: Weakly-supervised Temporal Activity Localization and Classification
Do you have extracted features from every frame or do you have taken some limited frames of the videos.
I tried to run this code it's giving map as .43 and classification score as 7.5 why it is so?
Hello, I find the link you provide only contains the i3d features for ActivityNet1.2. I wonder if you can provide the UntrimmedNet feature for ActivityNet1.2?
Dear Sir/Madam, Thanks for the code. It is really helpful for me to understand the temporal action localization. However, I want to try from the extracting features step but cannot...
Hi, could you please tell me that is still 25 fps used when extracting features of ActivityNet?
Can you provide any script for visualizing the data from the video using annotations for THUMOS14 ?
could you plz release the more detailed code/processing of extracting feature.
Hi, I trained your pytorch code on activityNet v1.2 dataset. But I can only get the results as follows. [email protected]= 47, [email protected]= 44, [email protected]=40, [email protected]=37, [email protected]= 33....it is much lower...
i cant download features from your provided link, something broken. Could you show me another link?
I used my own small dataset to extract I3D features. Then use main.py to train. Overfitting happened in training set. Do you have any suggestions? 
