Enhancing Robustness of Implicit Neural Representations Against Weight Perturbations

TL;DR

This paper investigates the vulnerability of Implicit Neural Representations to weight perturbations and proposes a novel robust loss function to significantly improve their resilience, demonstrated through multi-modal reconstruction experiments.

Contribution

It is the first study to analyze INR robustness and introduces a new loss function to enhance stability against weight perturbations.

Findings

Up to 7.5 dB PSNR improvement under noisy conditions

Minor weight perturbations cause significant performance drops

Proposed method effectively enhances INR robustness across modalities

Abstract

Implicit Neural Representations (INRs) encode discrete signals in a continuous manner using neural networks, demonstrating significant value across various multimedia applications. However, the vulnerability of INRs presents a critical challenge for their real-world deployments, as the network weights might be subjected to unavoidable perturbations. In this work, we investigate the robustness of INRs for the first time and find that even minor perturbations can lead to substantial performance degradation in the quality of signal reconstruction. To mitigate this issue, we formulate the robustness problem in INRs by minimizing the difference between loss with and without weight perturbations. Furthermore, we derive a novel robust loss function to regulate the gradient of the reconstruction loss with respect to weights, thereby enhancing the robustness. Extensive experiments on reconstruction tasks across multiple modalities demonstrate that our method achieves up to a 7.5~dB improvement in peak signal-to-noise ratio (PSNR) values compared to original INRs under noisy conditions.