Low-Light Image Enhancement with Normalizing Flow

TL;DR

This paper introduces a normalizing flow-based model for low-light image enhancement, effectively capturing the complex one-to-many relationship between low-light and normally exposed images, resulting in improved brightness, reduced noise, and richer colors.

Contribution

The paper proposes a novel invertible network using normalizing flow to model the conditional distribution of normally exposed images given low-light inputs, addressing limitations of previous deterministic methods.

Findings

Achieves better quantitative results on benchmark datasets.

Produces images with improved brightness and color richness.

Reduces noise and artifacts compared to prior methods.

Abstract

To enhance low-light images to normally-exposed ones is highly ill-posed, namely that the mapping relationship between them is one-to-many. Previous works based on the pixel-wise reconstruction losses and deterministic processes fail to capture the complex conditional distribution of normally exposed images, which results in improper brightness, residual noise, and artifacts. In this paper, we investigate to model this one-to-many relationship via a proposed normalizing flow model. An invertible network that takes the low-light images/features as the condition and learns to map the distribution of normally exposed images into a Gaussian distribution. In this way, the conditional distribution of the normally exposed images can be well modeled, and the enhancement process, i.e., the other inference direction of the invertible network, is equivalent to being constrained by a loss function that better describes the manifold structure of natural images during the training. The experimental results on the existing benchmark datasets show our method achieves better quantitative and qualitative results, obtaining better-exposed illumination, less noise and artifact, and richer colors.