SSD-Edit demonstrates an advanced editing framework for images and videos, suggesting improved instruction-following capabilities.
We propose SSD-Edit, the first instruction-based image and video editing framework built on Mamba-2 diffusion. No prior work has applied any Mamba variant to instruction-following editing. SSD-Edit introduces Source-Instruction Dual-Stream SSD, Selective Edit Masking via State Divergence for automatic edit localization, and Temporal Edit Propagation for consistent video editing.
No takes yet. Share an insight, caveat, or question.
Hiroki Abe (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: