[Feature Request] Video Analysis Toolkit Enhancement

[Aaron617]

Required prerequisites

• I have searched the Issue Tracker and Discussions that this hasn't already been reported. (+1 or comment there if it has.)
• Consider asking first in a Discussion.

Motivation

The current video analysis lack the ability to process long-video

Solution

1. In test hour-long challenge challenge , the best-performing solution is to segment video. can use structured text to aggregate information.
2. Gemini-2 may be better at video analysis (https://medium.com/@samarrana407/google-video-analyzer-gemini-2-0-b150c6f500fb)
3. Other reference papers:
   internvideo2.5: https://github.com/OpenGVLab/InternVideo/tree/main/InternVideo2.5
   movie chat：https://arxiv.org/pdf/2307.16449

Alternatives

No response

Additional context

No response

[lqjack]

@Aaron617 can I take the task, I have work the issue in another repo. camel-ai/owl#252

[lqjack]

@Aaron617 the PR in this project : #1852

[fengju0213]

@Aaron617 can I take the task, I have work the issue in another repo. camel-ai/owl#252

thanks， @lqjack for the contribution！we will review it