I wonder the video frame

[VoyageWang]

Hi, thanks for your great work! When doing the grounding task, will we input all the video sequences into the LLM or just an image?

[SongOnWater]

https://deepwiki.com/GeWu-Lab/Crab

[CserDu]

Hi, thanks for your great work! When doing the grounding task, will we input all the video sequences into the LLM or just an image?

Hello, we input an audio-image pair for grounding task.