SAPE: Spatially-Adaptive Progressive Encoding for Neural Optimization

TL;DR

SAPE is a novel encoding scheme that adaptively exposes frequency components in input signals for MLPs, improving their ability to learn high-frequency functions across diverse domains without complex preprocessing.

Contribution

The paper introduces SAPE, a spatially adaptive progressive encoding method that enhances neural networks' capacity to learn high-frequency signals through a feedback-driven, gradual frequency unmasking process.

Findings

Improves learning of high-frequency signals in MLPs.

Effective across various tasks including 3D shape transfer.

Maintains training stability without domain-specific preprocessing.

Abstract

Multilayer-perceptrons (MLP) are known to struggle with learning functions of high-frequencies, and in particular cases with wide frequency bands. We present a spatially adaptive progressive encoding (SAPE) scheme for input signals of MLP networks, which enables them to better fit a wide range of frequencies without sacrificing training stability or requiring any domain specific preprocessing. SAPE gradually unmasks signal components with increasing frequencies as a function of time and space. The progressive exposure of frequencies is monitored by a feedback loop throughout the neural optimization process, allowing changes to propagate at different rates among local spatial portions of the signal space. We demonstrate the advantage of SAPE on a variety of domains and applications, including regression of low dimensional signals and images, representation learning of occupancy networks, and a geometric task of mesh transfer between 3D shapes.