SwitchCraft: Training-Free Multi-Event Video Generation with Attention Controls

TL;DR

SwitchCraft is a training-free framework that improves multi-event video generation by aligning attention with event prompts, enhancing scene clarity and temporal consistency without additional training.

Contribution

It introduces Event-Aligned Query Steering and Auto-Balance Strength Solver to address multi-event video generation challenges without training.

Findings

Significantly improves prompt alignment and event clarity.

Enhances scene consistency in multi-event videos.

Operates without additional training or fine-tuning.

Abstract

Recent advances in text-to-video diffusion models have enabled high-fidelity and temporally coherent videos synthesis. However, current models are predominantly optimized for single-event generation. When handling multi-event prompts, without explicit temporal grounding, such models often produce blended or collapsed scenes that break the intended narrative. To address this limitation, we present SwitchCraft, a training-free framework for multi-event video generation. Our key insight is that uniform prompt injection across time ignores the correspondence between events and frames. To this end, we introduce Event-Aligned Query Steering (EAQS), which steers frame-level attention to align with relevant event prompts. Furthermore, we propose Auto-Balance Strength Solver (ABSS), which adaptively balances steering strength to preserve temporal consistency and visual fidelity. Extensive experiments demonstrate that SwitchCraft substantially improves prompt alignment, event clarity, and scene consistency compared with existing baselines, offering a simple yet effective solution for multi-event video generation.