Open niaolele opened 1 year ago
Hello Niaolele
We don't actually use the .mat files. We use a .npy ground-truth file instead :). I will delete the mat files later. We generate the .npy (GT) file by using the script: https://github.com/marctimjen/Weakly-supervised-anomaly-detection/blob/main/data/create_gt/create_test_gt.py.
I hope this helps :). Feel free to ask more questions about the code.
Do you by any chance have the RTFM pretrained network (for the UCF-crime)? I can't seem to get them downloaded from the original RTFM repo.
Best regards.
Hello, I have checked your repository to implement the RTFM code. I wanted to try it on another custom dataset However, I would appreciate any tips of how to generate the .mat annotation files? could you provide the script used to generate those files to begin with?
Thank you!