Shadman Sakib

Running video_jpg_ucf101_hmdb51.py gives error. UnicodeDecodeError: 'utf-8' can't decode byte 0x8c in position 3864: invalid start byte

1

I'm trying to train the model on hmdb51 dataset. I downloaded everything and ran utils/video_jpg_ucf101_hmdb51.py. After running for a while and converting some videos into jpgs it gives the following...

After running generate_result_video.py I get the following errors: ` 1. ../videos\: No such file or directory 2. ValueError: could not convert string to float: 'er' ` Is there anything wrong...

Was the video classification example trained from scratch? Or were the weights from I3d module used?