search

wangbo-zhao / 2021CVPR-WSVSOD

The code for CVPR2021 Weakly Supervised Video Salient Object Detection
43 stars 4 forks source link

Question about convlstm. #3

Closed aflyingnoob closed 2 years ago

aflyingnoob commented 3 years ago

Sorry to disturb you, I have a question about lstm which is used by your paper, the video frames are large, even some video contains 400+ frames, if we send all of them to lstm, it would induce memory overflow. Or you cut the 400+ frames to several parts and send them to lstm separately? Thanks.

wangbo-zhao commented 2 years ago

Yes. To the best of my knowledge, most methods for video usually cut the video to several parts, especially methods for video segmentation.