Search papers, labs, and topics across Lattice.
The paper introduces Soap2Soap, a multi-agent framework for long-horizon video-to-video generation that remakes entire episodes or films while preserving narrative structure, motion, and character identity. It enforces long-term language-visual consistency using a Dual-Bridge Consistency mechanism with a scene-aware screenplay and dynamic visual reference anchors. The framework also incorporates batch keyframe consistency and a closed-loop verification agent to mitigate drift and ensure fidelity, achieving superior performance on the SoapBench dataset compared to existing commercial APIs.
Achieve series-level cinematic remaking with Soap2Soap, a multi-agent framework that maintains narrative fidelity and character consistency across hundreds of shots, outperforming commercial video generation APIs.
We study series-level cinematic remaking, a long-horizon video-to-video generation problem that localizes full episodes or films via stylization or actor replacement while strictly preserving narrative structure, motion choreography, and character identity across hundreds of shots. Existing video generation and editing pipelines often break down in this regime due to compounding identity drift, background mutation, and semantic erosion under large camera motions and viewpoint changes. We propose Soap2Soap, a multi-agent framework that enforces long-term language-visual consistency through a Dual-Bridge Consistency mechanism: a scene-aware JSON screenplay serving as a persistent semantic backbone, and dynamically allocated visual reference anchors at both scene and shot levels. To suppress drift before video synthesis, we introduce batch keyframe consistency, jointly generating multiple keyframes in a shared latent context via a grid-based formulation. A closed-loop verification agent further audits identity, stability, and alignment to trigger selective regeneration. Experiments on SoapBench demonstrate strong improvements over commercial video generation APIs in long-term consistency and narrative fidelity.
NUS