ReFiNe: Recursive Field Networks for Cross-modal Multi-scene Representation

TL;DR

ReFiNe introduces a recursive hierarchical neural field approach for efficient, high-precision multi-shape representation that outperforms existing methods in reconstruction and compression, with low memory requirements.

Contribution

The paper presents a novel recursive hierarchical formulation for neural fields that enhances multi-shape encoding accuracy and efficiency without auxiliary data structures.

Findings

Achieves state-of-the-art multi-scene reconstruction results.

Provides highly compressed shape latent spaces.

Supports continuous field queries for applications like raytracing.

Abstract

The common trade-offs of state-of-the-art methods for multi-shape representation (a single model "packing" multiple objects) involve trading modeling accuracy against memory and storage. We show how to encode multiple shapes represented as continuous neural fields with a higher degree of precision than previously possible and with low memory usage. Key to our approach is a recursive hierarchical formulation that exploits object self-similarity, leading to a highly compressed and efficient shape latent space. Thanks to the recursive formulation, our method supports spatial and global-to-local latent feature fusion without needing to initialize and maintain auxiliary data structures, while still allowing for continuous field queries to enable applications such as raytracing. In experiments on a set of diverse datasets, we provide compelling qualitative results and demonstrate state-of-the-art multi-scene reconstruction and compression results with a single network per dataset.