English

Improved Techniques for Learning to Dehaze and Beyond: A Collective Study

Computer Vision and Pattern Recognition 2018-07-31 v2 Machine Learning

Abstract

Here we explore two related but important tasks based on the recently released REalistic Single Image DEhazing (RESIDE) benchmark dataset: (i) single image dehazing as a low-level image restoration problem; and (ii) high-level visual understanding (e.g., object detection) of hazy images. For the first task, we investigated a variety of loss functions and show that perception-driven loss significantly improves dehazing performance. In the second task, we provide multiple solutions including using advanced modules in the dehazing-detection cascade and domain-adaptive object detectors. In both tasks, our proposed solutions significantly improve performance. GitHub repository URL is: https://github.com/guanlongzhao/dehaze

Keywords

Cite

@article{arxiv.1807.00202,
  title  = {Improved Techniques for Learning to Dehaze and Beyond: A Collective Study},
  author = {Yu Liu and Guanlong Zhao and Boyuan Gong and Yang Li and Ritu Raj and Niraj Goel and Satya Kesav and Sandeep Gottimukkala and Zhangyang Wang and Wenqi Ren and Dacheng Tao},
  journal= {arXiv preprint arXiv:1807.00202},
  year   = {2018}
}

Comments

updated: typo fixed and some other improvements on writing

R2 v1 2026-06-23T02:46:59.373Z