Auditing Training-Free 3D Shape Retrieval with Diffused Geodesic Moments
For researchers in 3D shape retrieval, this work provides a reproducible protocol-cascade analysis and recommendations for designing and reporting training-free descriptors, highlighting that input field and aggregation dominate moment formula.
This paper audits training-free 3D shape retrieval by isolating protocol effects (normalization, aggregation, etc.) and introduces Diffused Geodesic Moments (DGM) as a baseline. Results show that aggregation-matched experiments reveal Geometric Moment Shape Descriptor (GMSD-HKS) achieves highest mAP (0.621/0.820 on FAUST-Reg, 0.865/0.963 on TOSCA), while DGM is useful only when sparse solves or symmetry-informative seeds are priorities.
Reported retrieval scores for training-free shape descriptors conflate local signal design, normalization, aggregation, codebook fitting, and metric choices, making isolated component evaluation difficult. This paper reframes descriptor evaluation as a {\em protocol audit}. We introduce Diffused Geodesic Moments (DGM), a seed-conditioned descriptor that computes sparse implicit heat responses, converts them to distance-like fields, and summarizes each vertex by low-order moments across seeds and scales. DGM is used both as a practical non-spectral baseline and as an instrument for isolating protocol effects. On the registered FAUST benchmark split (FAUST-Reg) and the TOSCA shape collection, aggregation-matched experiments show that an independent Geometric Moment Shape Descriptor baseline built on Heat Kernel Signature features (GMSD-HKS) obtains the highest scores in this implementation ($0.621/0.820$ and $0.865/0.963$ mean average precision (mAP)/top-1), Wave Kernel Signature (WKS) remains a strong classical signal, and DGM is useful mainly when sparse solves, non-spectral deployment, or symmetry-informative seed frames are priorities. The broader finding is methodological: the input field and aggregation protocol can dominate the moment formula. The paper contributes a reproducible protocol-cascade analysis, a cross-shape alignment diagnostic for functional-map compatibility, and concrete recommendations for designing and reporting training-free shape descriptors.