Closed hanah-fei closed 4 years ago
Average length of gestures in jester dataset is around 34. So, middle 32 frames of gestures are used with downsampling of 2 (taking every sexond frame) in order to calculate the video accuracy. I hope this explains your question.
@hanah-fei hello, did you train the jester dataset successfully?
Thanks for your nice work! But I didn't figure out how to calculate the accuracy of the jester dataset, can you explain it? Thanks a lot.