, optical flow, point tracks), and video manipulation.
A single video analogy framework unifies diverse tasks and generalizes to unseen scenarios
ViGeo uses spatiotemporal canvas completion to perform one-shot learning across camera control, modal-guided generation, and video manipulation.
Academic
Chia-Hsiang Kao · Belinda Zeng · Bharath Hariharan · Menglin Jia
Cornell University · Meta
Research Digest··3 min read
The authors introduce ViGeo, a framework that extends visual analogy to video by treating all tasks as spatiotemporal canvas completion.
Why this paper
From Cornell University and Meta
In one line
ViGeo extends visual analogy to video by training a single model to complete spatiotemporal canvases, unifying diverse video tasks without task-specific pipelines.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ·No compute details found
- ·No stated limitations found
- ·No benchmark numbers found
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§