World-R1: Reinforcing 3D Constraints for Text-to-Video Generation
For text-to-video generation, this work addresses the bottleneck of geometric inconsistency by introducing a scalable RL-based alignment method, though it is incremental over existing 3D prior injection approaches.
World-R1 uses reinforcement learning to align text-to-video generation with 3D constraints, achieving enhanced geometric consistency without architectural changes. The method significantly improves 3D coherence while maintaining visual quality.
Recent video foundation models demonstrate impressive visual synthesis but frequently suffer from geometric inconsistencies. While existing methods attempt to inject 3D priors via architectural modifications, they often incur high computational costs and limit scalability. We propose World-R1, a framework that aligns video generation with 3D constraints through reinforcement learning. To facilitate this alignment, we introduce a specialized pure text dataset tailored for world simulation. Utilizing Flow-GRPO, we optimize the model using feedback from pre-trained 3D foundation models and vision-language models to enforce structural coherence without altering the underlying architecture. We further employ a periodic decoupled training strategy to balance rigid geometric consistency with dynamic scene fluidity. Extensive evaluations reveal that our approach significantly enhances 3D consistency while preserving the original visual quality of the foundation model, effectively bridging the gap between video generation and scalable world simulation.