Refine Now, Query Fast: A Decoupled Refinement Paradigm for Implicit Neural Fields

Tianyu Xiong, Skylar Wurster, Han Wei Shen

Keywords: implicit neural representation, scene representation network, ensemble simulation, scientific simulation

TL;DR: We propose a "Refine Now, Query Fast" paradigm for INR surrogates, boosting representation fidelity with deep expressive networks at the high inference speed of embedding-based architectures.

Abstract: Implicit Neural Representations (INRs) have emerged as promising surrogates for large 3D scientific simulations due to their ability to continuously model spatial and conditional fields, yet they face a critical fidelity-speed dilemma: deep MLPs suffer from high inference cost, while efficient embedding-based models lack sufficient expressiveness. To resolve this, we propose the Decoupled Representation Refinement (DRR) architectural paradigm. DRR leverages a deep refiner network, alongside non-parametric transformations, in a one-time offline process to encode rich representations into a compact and efficient embedding structure. This approach decouples slow neural networks with high representational capacity from the fast inference path. We introduce DRR-Net, a simple network that validates this paradigm, and a novel data augmentation strategy, Variational Pairs (VP) for improving INRs under complex tasks like high-dimensional surrogate modeling. Experiments on several ensemble simulation datasets demonstrate that our approach achieves state-of-the-art fidelity, while being up to 27$\times$ faster at inference than high-fidelity baselines and remaining competitive with the fastest models. The DRR paradigm offers an effective strategy for building powerful and practical neural field surrogates and INRs in broader applications, with a minimal compromise between speed and quality.

Primary Area: applications to physical sciences (physics, chemistry, biology, etc.)

Submission Number: 8494