RESEARCH PAPER

Dream2Flow: Bridging Video Generation and Open-World Manipulation with 3D Object Flow

Karthik Dharmarajan; Wenlong Huang; Jiajun Wu; Li Fei-Fei; Ruohan Zhang

Classification

View four quadrants
Major category
WAMs
Architecture
Pending verification
Prediction paradigm
Other mechanisms
Source review status
Not assigned

Category review. Generated video futures become object-level 3D flow references, which model-based optimization or RL converts into executable robot behavior. The control solver is not automatically a learned IDM. The described action mechanism is external planning, model-assisted policy optimization, geometric tracking, or video-conditioned control; the source does not establish joint future/action generation or an IDM action decoder. Reading evidence

AT A GLANCE

Contribution

Dream2Flow converts generated human-interaction videos into 3D object trajectories, then uses domain-specific optimization or reinforcement learning to make a robot realize them. Its central benefit is an object-level interface across embodiments; its reliability still depends on video geometry, tracking, contact assumptions and the downstream controller.

Abstract

An abstract has not been added yet.

Affiliations

Stanford University