PolyU's OmniColor unifies multi-signal lineart colorization in an ECCV 2026 paper
jiqizhixin · x · 2026-09-05
Hong Kong Polytechnic University and Sichuan University present OmniColor, an ECCV 2026 framework for unified multi-modal lineart colorization. Animators juggle conflicting signals — text descriptions, sparse color scribbles, character reference sheets, recent frames — and existing methods handle at most one or two. OmniColor splits inputs into spatially-aligned conditions (lineart, color hints, frames) needing pixel-level precision and semantic reference conditions (text, character images), each with its own encoding strategy, training objective and redundancy elimination, combined via an adaptive gating module to support any combination of control signals.
More from Multimodal
- AI video now pulls off orbiting, diving camera moves around a speeding bicycle — umesh_ai · 2026-10-23
- AI art showcase: gothic winged warrior "Azrael-42" with porcelain-crack armor — LudovicCreator · 2026-10-23
- Moody AI-reimagined Harry Potter portraits showcase stunning cinematic generation — azed_ai · 2026-10-23
- Claude Code + ComfyUI MCP workflow for client work: paid models or open weights? — newaccount47 · 2026-10-23
- AI-generated character sheet wows: melancholy traveling cat 'Finch' fully spec'd — azed_ai · 2026-10-23
- Viral prompt template turns AI image models into children's book character-sheet generators — azed_ai · 2026-10-23