I downloaded the HOPE dataset and ran train.py without modifying the CleanVisiiDopeLoader data processing class code. The result was an error about the missing 'visibility' keyword. I confirmed that the JSON files in the downloaded hope_video dataset indeed do not contain the 'visibility' keyword. So is this because I didn't modify the dataset format? Or did I make a mistake in some other steps?
I downloaded the HOPE dataset and ran train.py without modifying the CleanVisiiDopeLoader data processing class code. The result was an error about the missing 'visibility' keyword. I confirmed that the JSON files in the downloaded hope_video dataset indeed do not contain the 'visibility' keyword. So is this because I didn't modify the dataset format? Or did I make a mistake in some other steps?