- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 3
- Soundness
- 2
Read review
Paper summary
The paper proposes 3D-Balanced-R1 for video generation rl. Beyond World-R1, it mainly introduces i) visibility-based reconstruction reward re-weighting; ii) reward renorm based on optimal transport and iii) early-late two stage policy updates. Experiments on two wan dit backbones show improved 3d score and camera following, and better training efficiency.
Critical strengths
- The proposed approaches are intuitive and interpretable, with experiment results supporting the effectiveness.
- The comparison with LTX-2.5-22B is a welcome addition.
- The presentation, including the figure visualization and math formulation, is concise and clear, making it easy for readers to follow.
Critical weaknesses
- The methodological contribution is somewhat incremental. World-R1 already discusses the near-static hacking problem and some mitigation approaches in Appendix D.7. Therefore, the main contribution here is the proposed visibility correction rather than identifying the issue. In addition, given that World-R1 has discussed the mitigations, the authors are suggested to present the potential issues with World-R1’s method and explain why the proposed modeling of S_vis is superior.
- The videos in the supplementary show visual trade-offs that are not sufficiently discussed. In “A steampunk city square,” the proposed method successfully follows “Camera orbit right,” but the requested “richly printed poster illustration” style is less pronounced than in the baseline. Plus, in “silk and tile” and “rail terminal”, the visual quality also appears noticeably worse. We know that rl can lead to similar de-stylization issues, but ‘ours’ appears more severe than World-R1.
- The early-late ablation compares the proposed two-step update with a single-step baseline, which may be unfair since the number of steps differs. It would be more appropriate to compare {early, late} v.s. {random, random} under the same two-step sampling budget (here random mean randomly sample a step in all possible timesteps)
- (minor) The timestep notation is inconsistent: Section 3 defines t in[0,1], whereas Section 4.3 uses t=T-s as an integer transition index.
Questions for authors
- Can S_vis distinguish the reduced overlap caused by camera movement from that caused by geometric inconsistency (for example, a relatively large object moves or just disappears inexplicably)? If inconsistent frames also receive a visibility bonus, could this weaken the reconstruction reward’s ability to penalize them?