University of Leicester
Browse
AAAI2022.pdf (4.68 MB)

Semantically Contrastive Learning for Low-light Image Enhancement

Download (4.68 MB)
conference contribution
posted on 2022-01-13, 14:31 authored by D Liang, L Li, M Wei, S Yang, L Zhang, W Yang, Y Du, Huiyu Zhou

Low-light image enhancement (LLE) remains challenging

due to the unfavorable prevailing low-contrast and weakvisibility problems of single RGB images. In this paper, we

respond to the intriguing learning-related question – if leveraging both accessible unpaired over/underexposed images

and high-level semantic guidance, can improve the performance of cutting-edge LLE models? Here, we propose an effective semantically contrastive learning paradigm for LLE

(namely SCL-LLE). Beyond the existing LLE wisdom, it

casts the image enhancement task as multi-task joint learning,

where LLE is converted into three constraints of contrastive

learning, semantic brightness consistency, and feature preservation for simultaneously ensuring the exposure, texture, and

color consistency. SCL-LLE allows the LLE model to learn

from unpaired positives (normal-light)/negatives (over/underexposed), and enables it to interact with the scene semantics to regularize the image enhancement network, yet the

interaction of high-level semantic knowledge and the lowlevel signal prior is seldom investigated in previous methods. Training on readily available open data, extensive experiments demonstrate that our method surpasses the state-of-thearts LLE models over six independent cross-scenes datasets.

Moreover, SCL-LLE’s potential to beneft the downstream semantic segmentation under extremely dark conditions is discussed. Source Code: https://github.com/LingLIx/SCL-LLE.


History

Author affiliation

School of Computing and Mathematical Sciences, University of Leicester

Source

Thirty-Sixth AAAI Conference on Artificial Intelligence (AAAI-22), February 22 - March 1, 2022.

Version

  • AM (Accepted Manuscript)

Published in

Proceedings of the AAAI Conference on Artificial Intelligence 2022

Volume

36

Issue

No. 2: AAAI-22 Technical Tracks 2

Pagination

1555-1563

Publisher

Association for the Advancement of Artificial Intelligence

Acceptance date

2021-12-01

Copyright date

2022

Available date

2022-03-01

Language

en

Usage metrics

    University of Leicester Publications

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC