Video Generation Workflow Using Depth Maps and Audio

PurzBeats · x · 2026-07-19

The author shares a video generation workflow: first, convert a reference video into a depth map, then feed it into Seedance 2.0 as a reference video.

Next, add a reference image, set the duration, resolution, and aspect ratio, passthrough the audio from the reference video into the final result, and simply run it to generate the output.

Related event: Seedance 2.0 in Action: Advanced Workflows and Model Selection(7 posts)→

Original post →

More from Multimodal

Multimodal channel →