Saved in:
Bibliographic Details
Main Authors: Kao, Chih-Chen, Makowski, Grzegorz, Fujieda, Shin, Harada, Takahiro
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2604.24666
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908995695935488
author Kao, Chih-Chen
Makowski, Grzegorz
Fujieda, Shin
Harada, Takahiro
author_facet Kao, Chih-Chen
Makowski, Grzegorz
Fujieda, Shin
Harada, Takahiro
contents We extend the Locally-Subdivided Neural Intersection Function (LSNIF) to support parameterized deformable and animated geometry. Our approach introduces a rest-space and deformed-space formulation inspired by meshless rendering, allowing ray samples to be mapped back to a canonical space where a single neural network represents geometry consistently across poses without retraining. To maintain accuracy under deformation-aware training, we incorporate scale-invariant distance regression, uncertainty-weighted multi-task learning, and a hybrid positional-grid encoding. The resulting method preserves the compactness and efficiency of LSNIF while enabling robust neural intersection prediction for dynamic geometry.
format Preprint
id arxiv_https___arxiv_org_abs_2604_24666
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Voxel Deformation-Aware Neural Intersection Function
Kao, Chih-Chen
Makowski, Grzegorz
Fujieda, Shin
Harada, Takahiro
Graphics
We extend the Locally-Subdivided Neural Intersection Function (LSNIF) to support parameterized deformable and animated geometry. Our approach introduces a rest-space and deformed-space formulation inspired by meshless rendering, allowing ray samples to be mapped back to a canonical space where a single neural network represents geometry consistently across poses without retraining. To maintain accuracy under deformation-aware training, we incorporate scale-invariant distance regression, uncertainty-weighted multi-task learning, and a hybrid positional-grid encoding. The resulting method preserves the compactness and efficiency of LSNIF while enabling robust neural intersection prediction for dynamic geometry.
title Voxel Deformation-Aware Neural Intersection Function
topic Graphics
url https://arxiv.org/abs/2604.24666