A depth iterative illumination estimation network for low-light image enhancement based on retinex theory

Article Engineering, Electrical & Electronic

RetinexDIP: A Unified Deep Framework for Low-Light Image Enhancement

Zunjin Zhao et al.

Summary: This paper proposes a novel generative strategy and a unified deep framework for enhancing low-light images. Experimental results demonstrate the superiority of the proposed method.

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY (2022)

Add to Collection

Article Engineering, Electrical & Electronic

Diverse image enhancer for complex underexposed image

Ziaur Rahman et al.

Summary: The visual appearance of images changes with varying environmental light conditions. Adjusting the exposure of various distorted images is a complex process. Existing approaches have not achieved visually pleasing results or are limited to certain types of images. To address exposure in various distorted images, a diverse image enhancement model was proposed, improving brightness, contrast, color, and eliminating haze.

JOURNAL OF ELECTRONIC IMAGING (2022)

Add to Collection

Article Physics, Multidisciplinary

Low Light Image Enhancement Algorithm Based on Detail Prediction and Attention Mechanism

Yanming Hui et al.

Summary: Most LLIE algorithms only focus on enhancing the brightness of the image, neglecting the extraction of image details and leading to image distortion. This paper proposes the DELLIE algorithm, which utilizes deep learning to extract and fuse image detail features. Experimental results demonstrate that the DELLIE algorithm can effectively extract and recover image detail information while improving the luminance.

ENTROPY (2022)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

URetinex-Net: Retinex-based Deep Unfolding Network for Low-light Image Enhancement

Wenhui Wu et al.

Summary: In this paper, we propose a Retinex-based deep unfolding network (URetinex-Net) to enhance low-light images. By unfolding an optimization problem into a learnable network, the image can be decomposed into reflectance and illumination layers. By formulating the decomposition problem as an implicit priors regularized model, three learning-based modules are designed for data-dependent initialization, high-efficient optimization, and user-specified illumination enhancement. Experimental results demonstrate the effectiveness and superiority of the proposed method in low-light image enhancement.

2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022) (2022)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

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

Long Ma et al.

Summary: This paper proposes a new Self-Calibrated Illumination (SCI) learning framework for enhancing image brightness in real-world low-light scenarios. The framework achieves a balance between computational efficiency and visual quality through cascaded illumination learning, weight sharing, and a self-calibrated module.

2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2022) (2022)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

AdaViT: Adaptive Vision Transformers for Efficient Image Recognition

Lingchen Meng et al.

Summary: This paper introduces AdaViT, an adaptive computation framework that learns to derive usage policies to improve the inference efficiency of vision transformers for image recognition. Experimental results demonstrate that AdaViT achieves more than 2x improvement in efficiency without sacrificing accuracy.

2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR) (2022)

Add to Collection

Article Computer Science, Software Engineering

Structure revealing of low-light images using wavelet transform based on fractional-order denoising and multiscale decomposition

Ziaur Rahman et al.

Summary: In this study, a new model is proposed to enhance the visual quality of low-light images by preventing overenhancement, handling uneven illumination, and suppressing noise. Experimental results demonstrate that the proposed method achieves high efficacy and outperforms traditional approaches in terms of overall performance.

VISUAL COMPUTER (2021)

Add to Collection

Article Physics, Multidisciplinary

Retinex-Based Fast Algorithm for Low-Light Image Enhancement

Shouxin Liu et al.

Summary: The proposed Retinex-based fast algorithm is able to enhance low-light images by converting, gamma correction, Retinex modeling, and color saturation correction. The enhanced images generated from this algorithm exhibit better quality and lower computational complexity compared to other state-of-the-art methods.

ENTROPY (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

Adaptive Unfolding Total Variation Network for Low-Light Image Enhancement

Chuanjun Zheng et al.

Summary: The study proposes an adaptive unfolding total variation network (UTVNet) to enhance real low-light images by approximating noise level and learning noise level map, leading to superior performance over state-of-the-art methods.

2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV 2021) (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence

Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement

Risheng Liu et al.

Summary: The study proposed a new method named RUAS, which combines Retinex rule and designs optimization processes to discover low-light prior architectures from a compact search space, constructing a lightweight yet effective image enhancement network.

2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021 (2021)

Add to Collection

Proceedings Paper Computer Science, Artificial Intelligence