Divide and Conquer: Decoupled Representation Alignment for Multimodal World Models
Authors:
- Junyuan XiaoJXJunyuan Xiao
- Dingkang LiangDLDingkang Liang
- Yixuan YeYYYixuan Ye
- Tongtong SuTSTongtong Su
- Wenming YangWYWenming Yang
- 9 September 2026
- book chapter
- Published by Springer Nature
- p. 433-451
Abstract
No abstract availableKeywords
All Related Versions
This publication has 25 references indexed in Scilit:
- AETHER: Geometric-Aware Unified World Modeling2025
- Learning 4D Embodied World Models2025
- From Slow Bidirectional to Fast Autoregressive Video Diffusion Models2025
- Video Depth Anything: Consistent Depth Estimation for Super-Long Videos2025
- DepthCrafter: Generating Consistent Long Depth Sequences for Open-world Videos2025
- VGGT: Visual Geometry Grounded Transformer2025
- Scalable Diffusion Models with Transformers2023
- High-Resolution Image Synthesis with Latent Diffusion Models2022
- The Unreasonable Effectiveness of Deep Features as a Perceptual Metric2018
- The Pascal Visual Object Classes (VOC) ChallengeInternational Journal of Computer Vision, 2009