Benchmarking Single-Image Dehazing and Beyond

By Boyi Li, Wenqi Ren, Dengpan Fu, Dacheng Tao, Dan Feng, Wenjun Zeng, Zhangyang Wang. IEEE Transactions on Image Processing, 2017

2,219 citations recorded by the source corpus.

We present a comprehensive study and evaluation of existing single-image dehazing algorithms, using a new large-scale benchmark consisting of both synthetic and real-world hazy images, called REalistic Single-Image DEhazing (RESIDE). RESIDE highlights diverse data sources and image contents, and is divided into five subsets, each serving different training or evaluation purposes. We further provide a rich variety of criteria for dehazing algorithm evaluation, ranging from full-reference metrics to no-reference metrics and to subjective evaluation, and the novel task-driven evaluation. Experiments on RESIDE shed light on the comparisons and limitations of the state-of-the-art dehazing algorithms, and suggest promising future directions.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/abf75b81b96ff2cd262086675deba55f907108db

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Contrastive Learning for Compact Single Image Dehazing (2021, methodology): “Method SOTS [27] Dense-Haze [1] NH-HAZE [2] # Param PSNR SSIM PSNR SSIM PSNR SSIM”
DeblurGAN-v2: Deblurring (Orders-of-Magnitude) Faster and Better (2019, methodology): “We ﬁt a Bradley-Terry model [3] to estimate the subjective score for each method so that they can be ranked, with the identical routine following the previous benchmark work [24, 26].”
An Underwater Image Enhancement Benchmark Dataset and Beyond (2019, background): “The pairwise comparisons have been proven to be more robust and consistent in subjective perception than individual rating [68], [69], though it is laborious and time-consuming.”
MAXIM: Multi-Axis MLP for Image Processing (2022, methodology): “The RESIDE dataset [45] contains two subsets: Indoor Training Set (ITS) which contains 13,990 hazy images generated from 1399 clean ones, and Outdoor Training Set (OTS) that consists of 313,950 hazing images synthesized from 8,970 haze-free outdoor scenes.”
ALL Snow Removed: Single Image Desnowing Algorithm Using Hierarchical Dual-tree Complex Wavelet Representation and Contradict Channel Loss (2021, methodology): “In the synthesis process, first, we apply the famous haze dataset called RESIDE [39] and follow the synthesized procedure in [2] to simulate the veiling”
GridDehazeNet: Attention-Based Multi-Scale Network for Image Dehazing (2019, methodology): “After data cleaning, the Outdoor Training Set (OTS) of RESIDE contains a total of 296695 hazy outdoor images, generated from 8477 clear images with β ∈ [0 .”
DEA-Net: Single Image Dehazing Based on Detail-Enhanced Convolution and Content-Guided Attention (2023, methodology): “REalistic Single Image DEhazing (RESIDE) [40] is a widely-used dataset, which contains five subsets: Indoor Training Set (ITS), Outdoor Training Set (OTS), Synthetic Objective Testing Set (SOTS), Real-world Task-driven Testing Set (RTTS), and Hybrid Subjective Testing Set (HSTS).”
Single Image Dehazing via Multi-scale Convolutional Neural Networks with Holistic Edges (2019, methodology): “Table 4 Average PSNR and SSIM of dehazed results on the RESIDE dataset (Li et al. 2018) using different configurations Depth network (Eigen et al.”
Single Image Deraining: A Comprehensive Benchmark Analysis (2019, methodology): “As pointed out by several recent works [41, 15, 42, 43], the performance of high-level computer vision tasks, such as object detection and recognition, will deteriorate in the presence of various sensory and environmental degradations.”
Artificial intelligence in the creative industries: a review (2020, background): “A comprehensive study and an evaluation of existing single-image dehazing CNNbased algorithms are reported in [269].”
Real-World Underwater Enhancement: Challenges, Benchmarks, and Solutions Under Natural Light (2019, background): “Actually, studies on haze removal [56], [57] also found”
Omni-Kernel Network for Image Restoration (2024, methodology): “We perform ablation studies by training OKNet-S on the RESIDE-Indoor (Li et al. 2018) dataset for 300 epochs and evaluating on SOTS-Indoor (Li et al. 2018).”
Image Restoration with Mean-Reverting Stochastic Differential Equations (2023, methodology): “We train the IR-SDE on the RESIDE (Li et al., 2018) Indoor Training Set (ITS) and test it on the Synthetic Objective Testing Set (SOTS).”
Image Dehazing Transformer with Transmission-Aware 3D Position Embedding (2022, methodology): “Following previous works [10,20,30], we use ITS and OTS subsets of RESIDE dataset [18] as the training datasets and conduct the evaluations on SOTS subset that contains 500 indoor and 500 outdoor hazy images.”
Advancing Image Understanding in Poor Visibility Environments: A Collective Benchmark Study (2019, methodology): “We also try three state-of-the-art dehazing approaches: (a) AOD-Net 7 2) Results and Analysis: We evaluate the object detection performance on the original hazy images of RESIDE RTTS set using Mask R-CNN.”
EnlightenGAN: Deep Light Enhancement Without Paired Supervision (2019, methodology): “The classiﬁcation results after light enhancement could be treated as an indirect measure on semantic information preservation, as [28, 47] suggested.”