Hello,
In this dataset, the defect classes are labelled with text (e.g. "TJ", "ZW", etc.). In track1 for video defect classification, the classes are labelled with IDs (0, 1,, etc.). Is there any relation between them? If yes, what is the mapping? It would be ideal if we could re-use a method from track1 with some modifications for track2. In this case, it would be helpful to know the mapping between class labels and class IDs.
thanks
Sorry, we do not want to show the mapping between class labels and class IDs.
Please do not use dataset of track1 on track2 competition or use dataset of track2 on track1 competition.
You can use any other public dataset.
ok, thank you for your reply.
Posted by: sthoduka @ April 15, 2022, 12:31 p.m.