Depth Video converts every source frame into a grayscale depth visualization while keeping motion coherent across the clip.
Upload one MP4 video. The current public route manages depth estimation and output settings automatically.
The source duration is rounded up to the next whole second before the per-second rate is applied.
Similar Models
Model overview
Convert an MP4 source into a frame-aligned grayscale depth visualization for compositing, scene understanding, reconstruction prep, and visual-effects workflows.
Upload a single MP4 clip and receive a matching video whose frames visualize estimated scene depth.
Depth estimates remain coherent as the camera and subjects move, reducing frame-to-frame flicker in downstream work.
Create a task with one public video URL, poll the shared task endpoint, or receive a best-effort webhook when processing finishes.
Model characteristics
Convert visible scene geometry into a grayscale near-to-far representation.
Keep the generated depth clip aligned with the source motion and frame sequence.
Use one required MP4 source while routing choices stay managed by the service.
FAQ
The current upload flow accepts one MP4 source video.
No. The current public route exposes only the source video and manages execution settings internally.
Poll the shared task endpoint or provide a top-level webhook URL when creating the task.
