Robust Network Learning via Inverse Scale Variational Sparsification

TL;DR

This paper introduces an inverse scale variational sparsification framework that enhances neural network robustness by preserving large-scale features while removing small-scale noise, improving resilience against diverse corruptions.

Contribution

It presents a novel inverse scale space approach that selectively retains large-scale features, offering a simple, efficient method to improve neural network robustness across multiple noise types.

Findings

Enhanced robustness against natural, adversarial, and low-resolution noise

Preservation of high-contrast details like textures and contours

Efficient integration into neural network training

Abstract

While neural networks have made significant strides in many AI tasks, they remain vulnerable to a range of noise types, including natural corruptions, adversarial noise, and low-resolution artifacts. Many existing approaches focus on enhancing robustness against specific noise types, limiting their adaptability to others. Previous studies have addressed general robustness by adopting a spectral perspective, which tends to blur crucial features like texture and object contours. Our proposed solution, however, introduces an inverse scale variational sparsification framework within a time-continuous inverse scale space formulation. This framework progressively learns finer-scale features by discerning variational differences between pixels, ultimately preserving only large-scale features in the smoothed image. Unlike frequency-based methods, our approach not only removes noise by smoothing small-scale features where corruptions often occur but also retains high-contrast details such as textures and object contours. Moreover, our framework offers simplicity and efficiency in implementation. By integrating this algorithm into neural network training, we guide the model to prioritize learning large-scale features. We show the efficacy of our approach through enhanced robustness against various noise types.