Run pretrained model on a video

[iMaTzzz]

First of all, thank you and congratulations for this amazing project !
Basically, I wanted to know if the code handles a video instead of datasets. Indeed, I'd like to try running the pretrained model on a 4K video to measure the inference time as well as the performance, to compare with other models. Thus, I wondered if this was possible and if so how ?

[AlexDotHam]

There is a solution in the previous issue Issue.

I hope this is useful for u

[iMaTzzz]

Thank you ! I was also trying the pytracking framework so I'll definitely try yours as well ! I'll let you know if that works