BNUDC: A Two-Branched Deep Neural Network for Restoring Images from Under-Display Cameras

Koh, Jaihyun; Lee, Jangho; Yoon, Sungroh

doi:10.1109/cvpr52688.2022.00199

Cited by 24 publications

(11 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This section describes our proposed DPFNet. Inspired by [9,35], we adopt a network structure with two independent branches to avoid the noise introduced by the mutual interference between the spatial and frequency domains during processing. DPFNet is mainly composed of two feature enhancement streams and a feature fusion.…”

Section: Methodsmentioning

confidence: 99%

DPFNet: A Dual-branch Dilated Network with Phase-aware Fourier Convolution for Low-light Image Enhancement

Zhuang¹,

Zheng²,

Lyu³

2022

Preprint

View full text Add to dashboard Cite

Low-light image enhancement is a classical computer vision problem aiming to recover normal-exposure images from low-light images. However, convolutional neural networks commonly used in this field are good at sampling low-frequency local structural features in the spatial domain, which leads to unclear texture details of the reconstructed images. To alleviate this problem, we propose a novel module using the Fourier coefficients, which can recover high-quality texture details under the constraint of semantics in the frequency phase and supplement the spatial domain. In addition, we design a simple and efficient module for the image spatial domain using dilated convolutions with different receptive fields to alleviate the loss of detail caused by frequent downsampling. We integrate the above parts into an end-to-end dual branch network and design a novel loss committee and an adaptive fusion module to guide the network to flexibly combine spatial and frequency domain features to generate more pleasing visual effects. Finally, we evaluate the proposed network on public benchmarks. Extensive experimental results show that our method outperforms many existing state-of-the-art ones, showing outstanding performance and potential. We release our code at https://github.com/Zhuangyunliang/DPFNet.

show abstract

Section: Methodsmentioning

confidence: 99%

DPFNet: A Dual-branch Dilated Network with Phase-aware Fourier Convolution for Low-light Image Enhancement

Zhuang¹,

Zheng²,

Lyu³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Koh et-al. [3] proposed a dual-branch network that handles high-frequency and low-frequency components separately. They introduced affine transformation connections to eliminate noise and preserve the structure of the image.…”

Section: Related Work a Udc Image Restorationmentioning

confidence: 99%

“…UDC IR methods. We select six state-of-the-art UDCIR methods for comparison, including DAGF [29], DISCNet [2], UDCUNet [4], BNUDC [3], SRUDC [7], and DWFormer [6]. We also include a general image restoration method called UFormer [58] for a comprehensive evaluation.…”

Section: Experiments a Implementation Detailsmentioning

confidence: 99%

‘Pre-training+Fine-tuning’ Model-based 1/2 Folded Image Restoration

Song,

Wei,

Lin

et al. 2023

2023 4th International Symposium on Computer Engineering and Intelligent Communications (ISCEIC)

View full text Add to dashboard Cite

This study delves into the enhancement of Under-Display Camera (UDC) image restoration models, focusing on their robustness against adversarial attacks. Despite its innovative approach to seamless display integration, UDC technology faces unique image degradation challenges exacerbated by the susceptibility to adversarial perturbations. Our research initially conducts an in-depth robustness evaluation of deep-learningbased UDC image restoration models by employing several whitebox and black-box attacking methods. This evaluation is pivotal in understanding the vulnerabilities of current UDC image restoration techniques. Following the assessment, we introduce a defense framework integrating adversarial purification with subsequent fine-tuning processes. First, our approach employs diffusion-based adversarial purification, effectively neutralizing adversarial perturbations. Then, we apply the fine-tuning methodologies to refine the image restoration models further, ensuring that the quality and fidelity of the restored images are maintained. The effectiveness of our proposed approach is validated through extensive experiments, showing marked improvements in resilience against typical adversarial attacks.

show abstract

“…[41]. Some methods design effective networks to predict normal-light images directly from low-light ones via a supervised learning strategy [2,27,15,45,18,53,4]. For instance, [45] learns a 3D lookup table, and [53] uses color consistency to constrain network training.…”

Section: Related Workmentioning

confidence: 99%

“…where ⊙ is point-wise multiplication. There still exist some other deep networks learning a mapping from lowlight images to normal-light ones, e.g., the supervised methods [2,27,15,45,18,53,4], and unsupervised methods [7,13,28]. However, the convolutional mechanism is usually limited by its inductive biases [30].…”

Section: Introductionmentioning

confidence: 99%

Low illumination image enhancement technology based on iterative dense residual network

Wang

Piao

Wang

et al. 2021

2021 International Conference on Electronic Information Engineering and Computer Science (EIECS)

View full text Add to dashboard Cite

This paper presents a novel network structure with illumination-aware gamma correction and complete image modelling to solve the low-light image enhancement problem. Low-light environments usually lead to less informative large-scale dark areas, directly learning deep representations from low-light images is insensitive to recovering normal illumination. We propose to integrate the effectiveness of gamma correction with the strong modelling capacities of deep networks, which enables the correction factor gamma to be learned in a coarse to elaborate manner via adaptively perceiving the deviated illumination. Because exponential operation introduces high computational complexity, we propose to use Taylor Series to approximate gamma correction, accelerating the training and inference speed. Dark areas usually occupy large scales in low-light images, common local modelling structures, e.g., CNN, SwinIR, are thus insufficient to recover accurate illumination across whole low-light images. We propose a novel Transformer block to completely simulate the dependencies of all pixels across images via a local-to-global hierarchical attention mechanism, so that dark areas could be inferred by borrowing the information from far informative regions in a highly effective manner. Extensive experiments on several benchmark datasets demonstrate that our approach outperforms state-of-the-art methods.

show abstract

BNUDC: A Two-Branched Deep Neural Network for Restoring Images from Under-Display Cameras

Cited by 24 publications

References 27 publications

DPFNet: A Dual-branch Dilated Network with Phase-aware Fourier Convolution for Low-light Image Enhancement

DPFNet: A Dual-branch Dilated Network with Phase-aware Fourier Convolution for Low-light Image Enhancement

‘Pre-training+Fine-tuning’ Model-based 1/2 Folded Image Restoration

Low illumination image enhancement technology based on iterative dense residual network

Contact Info

Product

Resources

About