More than 2 views

[bobbyohyeah]

Hello, first of all, fantastic work!

I’m curious if this would accept multiple viewpoints (i.e., multiple rooms in a building).

[hanyang-21]

Thank you for your interest and positive feedback! Currently, our baseline model, CogVideoX-v1.0, is limited to 49 frames per video, which makes it challenging to cover scenarios of multiple rooms. We encourage exploring with models like HunyuanVideo that support up to 200 frames per video. In future updates within the existing framework, we plan to incorporate multiple shorter videos to better represent more complex scenes with multiple viewpoints.