.npz score file for ucf101
ZealACMer opened this issue · 4 comments
Thank you very much for your prompt reply. The plan of mine is combine my method with the approach you proposed in this paper by late fusion. But I encountered some problem, which is the training list/test list of hmdb51 is different from the lists used in your paper, I wonder if there is a way to do late fusion with the hmdb51 .npz classification score file provided in this repository with the training list/ test list shown in https://github.com/chaoyuaw/pytorch-coviar/tree/master/data/datalists. Thank you so much for your help.
Hi,
I think you could refer to this page, and our renamed list come from ActionVLAD. Some details are as follows:
NOTE: For HMDB51, I renamed the videos to avoid issues with special characters in the filenames, and hence the numbers in the train/test files. The list of actual filenames is provided in data/hmdb51/train_test_lists/AllVideos.txt, and the new name for each video in that list is the 1-indexed line number of that video. The AllVideos_renamed.txt contains all the HMDB videos that are a part of one or all of the train/test splits (it has fewer entries than AllVideos.txt because some videos are not in any split). So, the video brush_hair/19 in that file (and in the train/test split files) would correspond to the line number 19 in AllVideos.txt.
Thank you so much for your prompt reply. Thanks a lot.