Unpaired Deep Image Dehazing Using Contrastive Disentanglement Learning

Chen, Xiang; Fan, Zhentao; Zheng, Zhuoran; Li, Yufeng; Huang, Yufeng; Dai, Longgang; Kong, Caihua; Li, Pengpeng

doi:10.1007/978-3-031-19790-1_38

Cited by 30 publications

(7 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To evaluate the generalization and effectiveness of our NSDNet on real smoke and hazy scenes, we compare with some state-of-the-art (SOTA) methods including DCP [19], DisentGAN [59], DAD [49], RefineNet [68], PSD [5], CDD-GAN [4] and D 4 [60]. For fair comparisons with these SOTA methods, we fine-tune them using our collected training datasets to achieve their best performance.…”

Section: Results On Real Smoke/hazy Datasetsmentioning

confidence: 99%

Non-aligned supervision for Real Image Dehazing

Fan¹,

Guo²,

Qian³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

Section: Results On Real Smoke/hazy Datasetsmentioning

confidence: 99%

Non-aligned supervision for Real Image Dehazing

Fan¹,

Guo²,

Qian³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…found that the blurred areas are mainly concentrated on the brightness channel of the YCrCb color space ( Wang et al., 2018 ) Therefore, it is possible to enhance the visual contrast of foggy scenes by recovering the missing texture information in the luminance channel. As for learning-based methods, the techniques such as attention ( Liu et al., 2019 ; Zhang et al., 2020 ), feature fusion ( Dong et al., 2020 ; Qin et al., 2020 ) and contrastive learning ( Wu et al., 2021 ; Chen et al., 2022 ) are widely used to improve single-image dehazing performance. Moreover, they outperform the traditional prior-based image dehazing methods.…”

Section: Related Workmentioning

confidence: 99%

All-in-one aerial image enhancement network for forest scenes

et al. 2023

View full text Add to dashboard Cite

Drone monitoring plays an irreplaceable and significant role in forest firefighting due to its characteristics of wide-range observation and real-time messaging. However, aerial images are often susceptible to different degradation problems before performing high-level visual tasks including but not limited to smoke detection, fire classification, and regional localization. Recently, the majority of image enhancement methods are centered around particular types of degradation, necessitating the memory unit to accommodate different models for distinct scenarios in practical applications. Furthermore, such a paradigm requires wasted computational and storage resources to determine the type of degradation, making it difficult to meet the real-time and lightweight requirements of real-world scenarios. In this paper, we propose an All-in-one Image Enhancement Network (AIENet) that can restore various degraded images in one network. Specifically, we design a new multi-scale receptive field image enhancement block, which can better reconstruct high-resolution details of target regions of different sizes. In particular, this plug-and-play module enables it to be embedded in any learning-based model. And it has better flexibility and generalization in practical applications. This paper takes three challenging image enhancement tasks encountered in drone monitoring as examples, whereby we conduct task-specific and all-in-one image enhancement experiments on a synthetic forest dataset. The results show that the proposed AIENet outperforms the state-of-the-art image enhancement algorithms quantitatively and qualitatively. Furthermore, extra experiments on high-level vision detection also show the promising performance of our method compared with some recent baselines.

show abstract

“…In recent years, contrastive learning has been introduced into low-level vision tasks such as image-to-image translation [54], deraining [55], [56], and dehazing [57], [58]. For instance, Chen et al [58] proposed an unsupervised contrastive CDD-GAN framework based on CycleGAN [59] for image dehazing, where positive and negative samples are sampled from the hazy domain and clear domain, respectively. Similarly, Ye et al [55] devised a novel non-local contrastive learning mechanism that leverages the inherent self-similarity property for image deraining.…”

Section: Contrastive Learningmentioning

confidence: 99%