Integrating perceptual cues with mixture-of-experts for low-light image restoration.

Abstract

Visual perception of nighttime images is often impaired by degradations: low-light, noise, motion blur, and low-resolution. While recent methods have made progress in jointly solving these degradations, the diversity of patterns and intensities in degradation has not been properly considered, leading to inconsistent illumination and unintended artifacts. In response, we propose to integrate perceptual cues with mixture-of-experts (IPCMoE) to achieve flexible processing for low-light low-quality images. By exploiting the perceptual cues, we strategically combine dedicated experts with the selective collaboration approach for feature enlightening and texture restoration. To this end, we develop perceptual-integrated MoEs by designing customized routers and task-dependent experts. Specifically, the texture memorial MoE is developed to preserve valuable features to restore high-fidelity details, and the enhancement MoE that adaptively integrates enlightening cues and texture cues is designed to formulate the relationship between feature enlightening and texture restoration, thereby achieving dynamic image processing. Compared to state-of-the-art models, our IPCMoE achieves superior performance on various benchmarks for handling complex low-light scenes.