Feature Denoising for Low-Light Instance Segmentation using Weighted Non-Local Blocks

About

This project proposed a novel generic block for denoising the feature maps of low-light images, which can be easily integrated into existing instance segmentation frameworks.

Funders

MyWorld Strength in Places

Abstract

Instance segmentation for low-light imagery remains largely unexplored due to the challenges imposed by such conditions, for example shot noise due to low photon count, color distortions and reduced contrast. In this paper, we propose an end-to-end solution to address this challenging task. Based on Mask R-CNN, our proposed method implements weighted non-local (NL) blocks in the feature extractor. This integration enables an inherent denoising process at the feature level. As a result, our method eliminates the need for aligned ground truth images during training, thus supporting training on real-world low-light datasets. We introduce additional learnable weights at each layer in order to enhance the network’s adaptability to real-world noise characteristics, which affect different feature scales in different ways.

Experimental results show that the proposed method outperforms the pretrained Mask R-CNN with an Average Precision (AP) improvement of +10.0, with the introduction of weighted NL Blocks further enhancing AP by +1.0.

Downloads

Code (on Github)

Code will be available soon.

Citation

@article{lin2024featdenoise,
    title={Feature Denoising for Low-Light Instance Segmentation using Weighted Non-Local Blocks},
    author={Lin, Joanne and Anatrasirichai, Nantheera and Bull, David},
    year={2024},
    publisher={arXiv}}