R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network

Hai, Jiang; Zhu, Xuan; Yang, Ren; Hao, Yutong; Zou, Fengzhu; Lin, Fang Cheng; Han, Songchen

doi:10.48550/arxiv.2106.14501

Cited by 4 publications

(3 citation statements)

References 51 publications

(45 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In qualitative evaluation, we tested on datasets from different scenarios, including the MIT dataset [46] and the LSRW dataset [47]. In the quantitative evaluation, we tested on the five datasets most commonly used for super resolution, including Set5, Set14, B100, Urban100, and Manga109.…”

Section: ) Datasets and Metricsmentioning

confidence: 99%

Semantic-Aware Guided Low-Light Image Super-Resolution

Ren,

Cao,

Tan

et al. 2024

IEEE Access

View full text Add to dashboard Cite

The single image super-resolution based on deep learning has achieved extraordinary performance. However, due to inevitable environmental or technological limitations, some images not only have low resolution but also low brightness. The existing super-resolution methods for restoring images through low-light input may encounter issues such as low brightness and many missing details. In this paper, we propose a semantic-aware guided low-light image super-resolution method. Initially, we present a semantic perception guided super-resolution framework that utilizes the rich semantic prior knowledge of the semantic network module. Through the semantic-aware guidance module, reference semantic features and target image features are fused in a quantitative attention manner, guiding low-light image features to maintain semantic consistency during the reconstruction process. Second, we design a self-calibrated light adjustment module to constrain the convergence consistency of each illumination estimation block by selfcalibrated block, improving the stability and robustness of output brightness enhancement features. Third, we design a lightweight super resolution module based on spatial and channel reconstruction convolution, which uses the attention module to further enhances the super-resolution reconstruction capability. Our proposed model surpasses methods such as RDN, RCAN, and NLSN in both qualitative and quantitative analysis of low-light image super-resolution reconstruction. The experiment proves the efficiency and effectiveness of our method.

show abstract

Section: ) Datasets and Metricsmentioning

confidence: 99%

Semantic-Aware Guided Low-Light Image Super-Resolution

Ren,

Cao,

Tan

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The illumination loss includes the perceptual loss, SSIM loss, and L1 loss. VGG-16 [33] is a classic deep convolutional neural network that has been trained on large-scale image classification tasks. Using VGG-16 as a loss function enables network learning to generate enhanced images with higher perceptual quality.…”

Section: Irecovery-netmentioning

confidence: 99%

DAF‐Retinex: Preserve the image detailed features and restore the reflected image

Huang,

Gao,

Wang

et al. 2024

IET Image Processing

View full text Add to dashboard Cite

Currently, deep learning methods for low‐light image enhancement tasks mainly focus on the illumination of images, while neglecting the problems of image noise and feature loss. To address this issue, this paper proposes a novel low‐light image enhancement network called DAF‐Retinex, based on the Retinex‐Net. To address the issue of image noise, different from traditional image denoise methods, this paper utilizes a fully convolutional neural network to denoise the reflection component, additionally, a denoising loss function is introduced to suppress noise. For preserving image details and extracting features, this paper creatively introduces self‐calibrated convolutions into low‐light image enhancement tasks, furthermore, a feature augmented attention block consisting of feature‐guided attention (FGA) is designed for feature learning to effectively enhance image illumination and extract image detail features. Experimental results demonstrate that the proposed algorithm in this paper effectively removes image noise and extracts detailed features, resulting in visually improved outcomes. On public datasets, the average improvement in objective evaluation metrics of image quality such as PSNR, SSIM, and NIQE are 1.13%, 4.12%, and 1.28%, respectively.

show abstract

“…Benchmarks Description and Metrics. As for low-light image enhancement, we randomly sampled 100 images from MIT dataset [2] and 50 testing image from LSRW dataset [9] for testing. We used two full-reference metrics including PSNR and SSIM, five no-reference metrics including DE [20], EME [1], LOE [23] and NIQE [23].…”

Section: Implementation Detailsmentioning

confidence: 99%

Toward Fast, Flexible, and Robust Low-Light Image Enhancement

Ma¹,

Ma²,

Liu³

et al. 2022

Preprint

View full text Add to dashboard Cite

ferent simple operations) and model-irrelevant generality (can be applied to illumination-based existing works to improve performance). Finally, plenty of experiments and ablation studies fully indicate our superiority in both quality and efficiency. Applications on low-light face detection and nighttime semantic segmentation fully reveal the latent practical values for SCI. The source code is available at https://github.com/vis-opt-group/SCI.

show abstract

R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network

Cited by 4 publications

References 51 publications

Semantic-Aware Guided Low-Light Image Super-Resolution

Semantic-Aware Guided Low-Light Image Super-Resolution

DAF‐Retinex: Preserve the image detailed features and restore the reflected image

Toward Fast, Flexible, and Robust Low-Light Image Enhancement

Contact Info

Product

Resources

About