テクノロジー
ByteDance「Seedance 2.0」が提示する動画生成の終着点：クリエイティブの「ガチャ」は終わるか
2026年2月、動画生成AIの勢力図が決定的に塗り替えられようとしている。TikTokを運営するByteDanceが週末に突如として公開した最新の動画生成モデル「Seedance 2.0」は、画質向上のみならず、これまで […]
Term
Dual-branch Diffusion Transformer
Overview
最終更新: 2026年6月8日The Dual-branch Diffusion Transformer is a technical architecture introduced in Seedance 2.0. Unlike traditional methods that add audio in post-processing, this architecture generates both visual pixels and audio data in parallel within the diffusion process, ensuring millisecond-level synchronization between actions and sounds.
Mentioned Articles
1 件External Mentions
7 件- arXiv teasr: training-efficient any-step diffusion transformer for real-world image super-resolution
- arXiv AHA-WAM:Asynchronous Horizon-Adaptive World-Action Modeling with Observation-Guided Context Routing
- arXiv Style-CCL: Content-Preserving Style Transfer via Curriculum Continual Learning
- arXiv Diffusion Image Generation with Explicit Modeling of Data Manifold Geometry
- arXiv Disentangling Generation and Regression in Stochastic Interpolants for Controllable Image Restoration
- arXiv Dual-Diffusional Generative Fashion Recommendation
- arXiv Talker-T2AV: Joint Talking Audio-Video Generation with Autoregressive Diffusion Modeling