ContextMaster: Interactive Multi-Shot Video Creation via Fixed-Budget Sparse Context Routing

2026-08-05

AI Summary

This work addresses the challenge of enabling flexible switching among generation, reference-guided synthesis, and editing tasks in multi-shot video creation while maintaining consistent shared historical context and computational efficiency. To this end, the authors propose a unified interactive framework for multi-shot video generation, featuring a character-aware contextual representation that integrates reusable clean context states with a sparse routing mechanism under fixed computational budgets (termed ConstraintSink). The framework further employs a two-stage privileged context distillation strategy—combining consistency distillation and distribution matching—to balance sparse access efficiency with high-quality few-step inference. Experiments demonstrate that the method outperforms task-specific baselines across three core tasks, significantly improving both task completion rates and cross-shot consistency. User studies confirm its support for flexible creative workflows, achieving real-time inference at 16 FPS on a single GPU.

Abstract

Recent video models increasingly support generation, reference conditioning, and editing within a single model, yet typically expose them as separate operations over fixed inputs. Practical creation unfolds across multiple shots, requiring one model to generate from text, follow a reference, or edit source footage while maintaining shared history. We formalize this setting as interactive multi-shot video creation (IMVC) and introduce ContextMaster, a unified model with a role-aware context representation for these operations. An interactive model must retain access to an expanding history without allowing the context read cost at each denoising step to grow. ContextMaster combines reusable clean context states with fixed budget sparse context routing and uses ConstraintSink to keep task constraints visible. To address the dual challenges of sparse context access and inference with few denoising steps, we propose a two-stage privileged context distillation framework, which transfers full context behavior from a dense teacher through consistency distillation and then refines deployment rollouts with distribution matching. Experiments on the three primitive tasks demonstrate improved task fulfillment and consistency across shots over specialized baselines. User studies further validate flexibly composed workflows, while the model reaches 16 FPS on a single GPU.

Problem

• interactive multi-shot video creation
• context routing
• video generation
• reference conditioning
• video editing

Innovation

• interactive multi-shot video creation
• sparse context routing
• context distillation
• role-aware context representation
• ConstraintSink