Rendering Nighttime Image Via Cascaded Color and Brightness Compensation

TL;DR

This paper introduces a new neural network-based ISP method for nighttime images, supported by a high-resolution dataset, achieving superior visual quality and competitive challenge rankings.

Contribution

The paper presents a novel cascaded color and brightness compensation network and a high-resolution nighttime RAW-RGB dataset for improved nighttime image rendering.

Findings

Our method outperforms traditional ISP pipelines in visual quality.

Achieved second place in NTIRE 2022 Night Photography Rendering Challenge.

Developed a high-resolution dataset with expert annotations.

Abstract

Image signal processing (ISP) is crucial for camera imaging, and neural networks (NN) solutions are extensively deployed for daytime scenes. The lack of sufficient nighttime image dataset and insights on nighttime illumination characteristics poses a great challenge for high-quality rendering using existing NN ISPs. To tackle it, we first built a high-resolution nighttime RAW-RGB (NR2R) dataset with white balance and tone mapping annotated by expert professionals. Meanwhile, to best capture the characteristics of nighttime illumination light sources, we develop the CBUnet, a two-stage NN ISP to cascade the compensation of color and brightness attributes. Experiments show that our method has better visual quality compared to traditional ISP pipeline, and is ranked at the second place in the NTIRE 2022 Night Photography Rendering Challenge for two tracks by respective People's and Professional Photographer's choices. The code and relevant materials are avaiable on our website: https://njuvision.github.io/CBUnet.