search

Epiphqny / VisTR

[CVPR2021 Oral] End-to-End Video Instance Segmentation with Transformers
https://arxiv.org/abs/2011.14503
Apache License 2.0
740 stars 96 forks source link

bbox and mask #38

Closed CuckooHXY closed 3 years ago

CuckooHXY commented 3 years ago

I try to visualize the predictions,and find mask is not wholely in bounding box.Should it be like that?Or is my visualization wrong? 0

Epiphqny commented 3 years ago

Hi @CuckooHXY, the visualization is right, the mask is globally predicted, therefore we didn't constrain the area to be in the box.

OfryZarfaty commented 3 years ago

I try to visualize the predictions,and find mask is not wholely in bounding box.Should it be like that?Or is my visualization wrong? 0

Hi, can you refer how to visualize the results.json?

Epiphqny commented 3 years ago

Hi @OfryZarfaty, The visualization is correct, the mask is not constrained by the box area.

matan2455 commented 3 years ago

Hey @CuckooHXY, How did you manage to visualize the results? I ran the model on a new video and got RLE format masks for it, just couldn't find any guide on how to visualize the results.