A unified video model with role-aware context and fixed-budget sparse attention enables interactive multi-shot generation, reference conditioning, and editing at 16 FPS on a single GPU.
Proceedings of the SIGGRAPH Asia 2025 Conference Papers , pages=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
ContextMaster: Interactive Multi-Shot Video Creation via Fixed-Budget Sparse Context Routing
A unified video model with role-aware context and fixed-budget sparse attention enables interactive multi-shot generation, reference conditioning, and editing at 16 FPS on a single GPU.