hitachinsk / FGT

[ECCV 2022] Flow-Guided Transformer for Video Inpainting
https://hitachinsk.github.io/publication/2022-10-01-Flow-Guided-Transformer-for-Video-Inpainting
MIT License
311 stars 32 forks source link

Not inpainting beginning and end of videos #42

Closed smandava98 closed 1 year ago

smandava98 commented 1 year ago

H @hitachinsk . I was testing on a few (longer videos) and I found that the inpainting works well for the middle of the video but the object (despite having the necessary masks) is still present in the beginning and end of the video.

To process the longer videos, I have the step size to be 1/10th of the total number of frames and the neighbor_stride to be 10. Can you please explain how I can modify parameters to inpaint fully? I am not quite clear on how neighbor frames and step together play a role.