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

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: this https URL

• Publication year

  2018

• Venue

  arXiv.org

• Publication date

  2018-06-30

• Fields of study

  Computer Science, Environmental Science

• Identifiers
  arXiv 1807.00202
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Domain Adaptive Faster R-CNN for Object Detection in the Wild

  2018cited by this paper
• Densely Connected Pyramid Dehazing Network

  2018cited by this paper
• Focal Loss for Dense Object Detection

  2017influential reference
• Robust emotion recognition from low quality and low bit rate video: A deep learning approach

  2017cited by this paper
• When Image Denoising Meets High-Level Vision Tasks: A Deep Learning Approach

  2017cited by this paper
• Loss Functions for Image Restoration With Neural Networks

  2017influential reference
• An All-in-One Network for Dehazing and Beyond

  2017cited by this paper
• RESIDE: A Benchmark for Single Image Dehazing

  2017influential reference
• Mask R-CNN

  2017influential reference
• End-to-End United Video Dehazing and Detection

  2017cited by this paper
• AOD-Net: All-in-One Dehazing Network

  2017influential reference
• Benchmarking Single-Image Dehazing and Beyond

  2017cited by this paper
• Enhance Visual Recognition Under Adverse Conditions via Deep Networks

  2017cited by this paper
• DehazeNet: An End-to-End System for Single Image Haze Removal

  2016influential reference
• Single Image Dehazing via Multi-scale Convolutional Neural Networks

  2016influential reference
• Non-local Image Dehazing

  2016cited by this paper
• Single Image Haze Removal Using Dark Channel Prior

  2016influential reference
• Studying Very Low Resolution Recognition Using Deep Networks

  2016cited by this paper
• A Fast Single Image Haze Removal Algorithm Using Color Attenuation Prior

  2015cited by this paper
• SSD: Single Shot MultiBox Detector

  2015cited by this paper
• Deep Residual Learning for Image Recognition

  2015cited by this paper
• Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

  2015influential reference
• Unsupervised Domain Adaptation by Backpropagation

  2014cited by this paper
• Investigating Haze-Relevant Features in a Learning Framework for Image Dehazing

  2014influential reference
• A comprehensive evaluation of full reference image quality assessment algorithms

  2012cited by this paper
• Deep photo: model-based photograph enhancement and viewing

  2008cited by this paper
• Image quality assessment: from error visibility to structural similarity

  2004cited by this paper
• Vision and the Atmosphere

  2002cited by this paper
• Instant dehazing of images using polarization

  2001cited by this paper
• Enhancement of color images in poor visibility conditions

  2000cited by this paper
• Chromatic framework for vision in bad weather

  2000cited by this paper
• Optics of the atmosphere--Scattering by molecules and particles

  1978cited by this paper

• Efficient image dehazing with synergic expert modulation

  2026cites this paper
• WDMamba: When Wavelet Degradation Prior Meets Vision Mamba for Image Dehazing

  2025cites this paper
• Spatial Frequency Modulation Network for Efficient Image Dehazing

  2025cites this paper
• Scattering media removal under the harsh conditions using adaptive removal via mask for scatter

  2025cites this paper
• Alpha-DehazeNet: single image dehazing via RGBA haze modeling and adaptive learning

  2025cites this paper
• Research on Power Inspection Image Defogging Method Based on Multilevel Feature Fusion Network

  2025cites this paper
• Diffusion Models in Low-Level Vision: A Survey

  2024cites this paper
• Efficient Scene Recovery Using Luminous Flux Prior

  2024cites this paper
• Spatial-Frequency Adaptive Remote Sensing Image Dehazing With Mixture of Experts

  2024cites this paper
• Mutual Information-driven Triple Interaction Network for Efficient Image Dehazing

  2023cites this paper
• ARODNet: adaptive rain image enhancement object detection network for autonomous driving in adverse weather conditions

  2023cites this paper
• Uncertainty-Driven Dehazing Network

  2022cites this paper
• Effects of haze and dehazing on deep learning-based vision models

  2022cites this paper
• Defogging Target Detection based on YOLOv4 and Deblurganv2

  2022cites this paper
• A Comprehensive Benchmark Analysis of Single Image Deraining: Current Challenges and Future Perspectives

  2021cites this paper
• A GAN-Based Input-Size Flexibility Model for Single Image Dehazing

  2021influential citation
• R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network

  2021cites this paper
• Advanced Multiple Linear Regression Based Dark Channel Prior Applied on Dehazing Image and Generating Synthetic Haze

  2021influential citation
• Debiased Subjective Assessment of Real-World Image Enhancement

  2021cites this paper
• Video Dehazing Based on Convolutional Neural Network Driven Using Our Collected Dataset

  2020influential citation
• Learning Interleaved Cascade of Shrinkage Fields for Joint Image Dehazing and Denoising

  2020cites this paper
• A Survey on Video Dehazing Using Deep Learning

  2020influential citation
• Moiré Pattern Removal via Attentive Fractal Network

  2020cites this paper
• VDFlow: Joint Learning for Optical Flow and Video Deblurring

  2020cites this paper
• Color-Constrained Dehazing Model

  2020cites this paper
• A compensation textures dehazing method for water alike area

  2020cites this paper
• CP-Net: Channel Attention and Pixel Attention Network for Single Image Dehazing

  2020cites this paper
• Deep Retinex Network for Single Image Dehazing

  2020cites this paper
• Haze-robust image understanding via context-aware deep feature refinement

  2020cites this paper
• Object Detection in Hazy Environment Enhanced by Preprocessing Image Dataset with Synthetic Haze

  2020influential citation
• UG 2+ Track 2: A Collective Benchmark Effort for Evaluating and Advancing Image Understanding in Poor Visibility Environments

  2019cites this paper
• Single Image Deraining: A Comprehensive Benchmark Analysis

  2019cites this paper
• Improved Bilateral Filtering for a Gaussian Pyramid Structure-Based Image Enhancement Algorithm

  2019cites this paper
• Accurate Transmission Estimation for Removing Haze and Noise From a Single Image

  2019cites this paper
• Enhancement of Weakly Illuminated Images by Deep Fusion Networks

  2019cites this paper
• Advancing Image Understanding in Poor Visibility Environments: A Collective Benchmark Study

  2019cites this paper
• Single Image Dehazing via Multi-scale Convolutional Neural Networks with Holistic Edges

  2019cites this paper
• End-to-end image dehazing via dual principles of discriminativeness

  2019cites this paper
• EnlightenGAN: Deep Light Enhancement Without Paired Supervision

  2019cites this paper
• FAMED-Net: A Fast and Accurate Multi-Scale End-to-End Dehazing Network

  2019cites this paper
• All-In-One Underwater Image Enhancement using Domain-Adversarial Learning

  2019cites this paper
• Framework for Underwater Dataset Generation and Classification Towards Modeling Restoration

  2019cites this paper
• U-Finger: Multi-Scale Dilated Convolutional Network for Fingerprint Image Denoising and Inpainting

  2018cites this paper
• Multiple Linear Regression Haze-Removal Model Based on Dark Channel Prior

  2018cites this paper
• Benchmarking Single-Image Dehazing and Beyond

  2017cites this paper