In recent years, low-light images have special properties and have been extensively studied. As deep learning technology advances, attention modules are increasingly employed in low-light image enhancement. We propose Discrete Wavelet Transform based Attention Network for Low-light Image Enhancement, our proposed Color Recovery Module initially recovers the color of low-light image, whereas the Color Adjustment Module focuses on useful information to adjust the color information. Finally, the overall details of the image are fine-tuned using 2D Discrete Wavelet Transform. The attention module is combined with the 2D Discrete Wavelet Transform to fully acquire the image information and ensure the information interaction in order to recover the image structure and details, and finally obtain a visually good and clean image. A series of experiments demonstrate that our network exceeds the state-of-the-art low-light image enhancement network to some extent.
Low-light images are usually affected by problems such as low illumination, noise, and color distortion, resulting in unsatisfactory image enhancement. Low-light image enhancement aims to improve the quality and visibility of low-light images. In this paper, a gray image based low-light image enhancement deep learning method is proposed, which converts the low-light image to gray image and then extracts the important information in the image as branches to assist the main network. To achieve better enhancement performance, we proposed denoising center block (DC block), which focusing on color restoration and information preservation. The experimental results show that the method proposed in this paper outperforms the existing low-light image enhancement methods in both objective and subjective evaluation metrics, and can effectively improve the clarity, contrast, and color reproduction of low-light images.
In recent years, significant progress has been made in image dehazing, but most dehazing convolutional neural networks only learn from hazy images to the corresponding feature maps of clean images, ignoring the details of the images. In this paper, a new flatness loss function is proposed for single image dehazing, thereby improving the overall effect of dehazing results. This loss function allows the model to focus on the texture features of hazy images and clean images and supervises edge information by reducing the flatness difference between clean pixels and blurred pixels. The experiments on the benchmark dataset using the flatness loss function on the single image dehazing model show that this method can effectively improve the quantitative performance of the model.
Massive multiple-input multiple-output (MIMO) system is one of the wireless technologies with great research significance based on channel state information (CSI). As the complexity of the CSI matrix increases, CSI feedback faces many challenges and limitations. This paper proposes a neural network composed of the encoder-decoder framework, which can effectively compress and recover CSI. The network utilizes a channel information generation module and multi-scale feature extraction module to enhance the information acquisition ability of the CSI matrix. At the same time, it reduces the computational complexity of the network by using lightweight architecture. Experimental results show that the network outperforms other advanced deep-learning methods at multiple compression rates.
Image enhancement tasks have garnered significant attention in recent years, particularly regarding images captured in low-light conditions. Due to their inherent under-illumination characteristics, these images often suffer from challenges such as low contrast, low brightness, and noise that hinder their quality. With the recent advancements in deep learning, convolutional neural networks (CNNs) have demonstrated a high potential in addressing these challenges. In this paper, we propose a deep CNN-based approach that enhances color components a and b through Lab-space decomposition of images, utilizing a deep unsupervised dehazing network model as a baseline. Our approach considers both algorithmic and image decomposition, splitting the low-light image enhancement task into two parallel subtasks. Extensive experiments demonstrate that our proposed approach outperforms previous methods.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.