- Rating
- 1
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
The paper regularises neural UDF learning with a "vector heat diffusion" term applied only early in training. This term reduces to a Hessian energy normalised by the mean gradient norm. It helps the field converge at the start of training, but novelty over existing Hessian regularisers is limited and not discussed.
Critical strengths
The regularisation seems to stabilise training and improve convergence.
Critical weaknesses
-
The proposed regulariser reduces to Hessian regularisation.
- Proposition 2 cancels the normalisation: (Eq. 19), so the smoothness term is the squared Frobenius norm of the Hessian, up to a global factor (Eq. 21). Since depends on , this factor also changes the effective weight of the term during training. Why choose rather than simply with a constant ?
- Hessian regularisation is well studied: in DiGS [1], which is also annealed to zero during training; in NSH [2]; and the Monge–Ampère term in S²DF [3]. These works are cited in the related work, but the proposed term is never related to them. The authors should acknowledge this equivalence explicitly and compare against these terms under the same schedule.
-
The motivation for vector heat diffusion is unclear.
- In vector heat diffusion, the source is imposed on and diffused into the volume. Here, the diffused field is , which cannot match this source on : is undefined there for a UDF, and for a smooth attaining its minimum on (Eq. 14), , so the alignment term is constant. The boundary condition of the diffusion is therefore ill-posed. Together with the previous point, it is unclear what the vector heat diffusion formulation adds beyond a Hessian penalty.
-
Proposition 1 is weak.
- It assumes the symmetry of a UDF rather than proving it: that is double-sided around follows from the definition of a UDF ( near a smooth point ), but not from the stated hypotheses , . Why not state this directly in terms of the UDF definition, and why introduce a general lower bound instead of ?
- The invariance under follows immediately from the absolute value in Eq. 3, so the random-flip experiment in C.3.1 verifies a property that holds by construction.
-
Missing ablations. The paper should ablate , , , , the schedule, and the chosen hyperparameters.
-
Errors in references.
- For "Neural unsigned distance fields for implicit function learning." (l. 521), the following authors are given "Julian Chibane, Gerard Pons-Moll, et al." while the true authors are "Julian Chibane, Aymen Mir, Gerard Pons-Moll", missing Aymen Mir.
- For HSDF (l.643), the following authors are given "Li Wang, Weikai Chen, Xiaoxu Meng, Bo Yang, Jintao Li, Lin Gao, et al.", while the true authors are "Li Wang, Jie Yang, Weikai Chen, Xiaoxu Meng, Bo Yang, Jintao Li, Lin Gao", missing Jie Yang.
- For “Voronoi-assisted diffusion for computing unsigned distance fields from unoriented points” (l. 572), the arXiv entry (2510.12524) has since been retitled “Voronoi-Assisted Optimization for Diffusing Unsigned Distance Fields from Unoriented Points”, the reference should be updated.
Questions for authors
- Could Table 1 report metrics on the whole dataset? Baselines fail on simple garments (e.g. S²DF in Fig. 8), whereas [3] reports near-perfect F-scores for S²DF, LevelSetUDF and CAP-UDF on 20 MGN shapes. What is the input density, and was Algorithm 1 applied to S²DF's squared output?
- How is chosen? Is computed per batch, and is it detached from the graph?
- No runtime or memory usage is reported, although second-order derivatives are computed at every off-surface sample.