Filters








30 Hits in 6.3 sec

Single Image Dehazing via Multi-scale Convolutional Neural Networks [chapter]

Wenqi Ren, Si Liu, Hua Zhang, Jinshan Pan, Xiaochun Cao, Ming-Hsuan Yang
2016 Lecture Notes in Computer Science  
In this paper, we propose a multi-scale deep neural network for single-image dehazing by learning the mapping between hazy images and their corresponding transmission maps.  ...  To train the multi-scale deep network, we synthesize a dataset comprised of hazy images and corresponding transmission maps based on the NYU Depth dataset.  ...  (b) Proposed multi-scale convolutional neural network.  ... 
doi:10.1007/978-3-319-46475-6_10 fatcat:e5dndlpoyrhodgm6iwoscqfgmu

Rich Feature Distillation with Feature Affinity Module for Efficient Image Dehazing [article]

Sai Mitheran, Anushri Suresh, Nisha J. S., Varun P. Gopi
2022 arXiv   pre-print
This work introduces a simple, lightweight, and efficient framework for single-image haze removal, exploiting rich "dark-knowledge" information from a lightweight pre-trained super-resolution model via  ...  Single-image haze removal is a long-standing hurdle for computer vision applications.  ...  teacher network to distill multi-scale information to the student dehazing network.  ... 
arXiv:2207.11250v1 fatcat:zrdj7l3jdfhkvnisquli6gafcy

A Comprehensive Review on Image Dehazing

Jini Elsa Joseph, college of Engineering Chengannur
2020 International Journal of Engineering Research and  
The total number of convolutional neural networks is six layers, which are convolution layer, a slice layer, element-by-element operation layer, multi-scale convolution layer, max pool layer, and convolution  ...  The scene transmission map is estimated by a coarse-scale network that predicts a holistic transmission map based on the entire image and refines dehazed results locally by the fine-scale network.  ... 
doi:10.17577/ijertv9is060822 fatcat:dtwkffa5mzhermdcew5mobhvlq

Classification-driven Single Image Dehazing [article]

Yanting Pei, Yaping Huang, Xingyuan Zhang
2019 arXiv   pre-print
Most existing dehazing algorithms often use hand-crafted features or Convolutional Neural Networks (CNN)-based methods to generate clear images using pixel-level Mean Square Error (MSE) loss.  ...  Experimental results demonstrate that the proposed method outperforms many recent state-of-the-art single image dehazing methods in terms of image dehazing metrics and classification accuracy.  ...  [9] propose a multi-scale deep neural network for haze removal, and the network consists of a coarse-scale sub-network for a holistic transmission map and a fine-scale sub-network for local refinement  ... 
arXiv:1911.09389v1 fatcat:snbtcfcycfhgbcsegibhtjnu74

A Survey of Deep Learning-Based Image Restoration Methods for Enhancing Situational Awareness at Disaster Sites: The Cases of Rain, Snow and Haze

Sotiris Karavarsamis, Ioanna Gkika, Vasileios Gkitsas, Konstantinos Konstantoudakis, Dimitrios Zarpalas
2022 Sensors  
These image restoration methods are: (a) deraining; (b) desnowing; (c) dehazing ones.  ...  This survey article is concerned with the emergence of vision augmentation AI tools for enhancing the situational awareness of first responders (FRs) in rescue operations.  ...  grid network, feature fusion GDNet [124] 3 sub-processes, multi-scale grid network MSCNN [125] 2 nets: coarse-and fine-scale MSCNN-HE [126] 3 nets: coarse-, fine-scale and holistic edge guided EMRA-Net  ... 
doi:10.3390/s22134707 pmid:35808203 pmcid:PMC9269588 fatcat:d34wmmvjznfcxfmh3vkj5qlk44

PAD-Net: A Perception-Aided Single Image Dehazing Network [article]

Yu Liu, Guanlong Zhao
2018 arXiv   pre-print
end-to-end dehazing neural network (AOD-Net) that uses the ℓ_2 loss.  ...  In this work, we investigate the possibility of replacing the ℓ_2 loss with perceptually derived loss functions (SSIM, MS-SSIM, etc.) in training an end-to-end dehazing neural network.  ...  [9] exploits a multi-scale CNN (MSCNN) that predicts a coarse-scale holistic transmission map of the entire image and refines it locally.  ... 
arXiv:1805.03146v1 fatcat:2gliwtemnfb6nepdj5imbfj6dy

2021 Index IEEE Transactions on Image Processing Vol. 30

2021 IEEE Transactions on Image Processing  
., +, TIP 2021 4883-4893 Single Image Dehazing via Dual-Path Recurrent Network.  ...  Multi-Scale Single Image Dehazing Using Laplacian and Gaussian Pyra-Rabiei, H., +, TIP 2021 1453-1460 mids.  ... 
doi:10.1109/tip.2022.3142569 fatcat:z26yhwuecbgrnb2czhwjlf73qu

Holistic Attention-Fusion Adversarial Network for Single Image Defogging [article]

Wei Liu, Cheng Chen, Rui Jiang, Tao Lu, Zixiang Xiong
2022 arXiv   pre-print
To address these issues, we develop a novel generative adversarial network, called holistic attention-fusion adversarial network (HAAN), for single image defogging.  ...  HAAN is designed to exploit the self-similarity of texture and structure information by learning the holistic channel-spatial feature correlations between the foggy image with its several derived images  ...  [43] first proposed a disentangled dehazing network consists of three generators and a multi-scale discriminator to produce defogged results from the foggy images.  ... 
arXiv:2202.09553v1 fatcat:eo4zadfqgbhdbb66ioqh6gixty

Lightweight and Efficient Image Dehazing Network Guided by Transmission Estimation from Real-World Hazy Scenes

Zhan Li, Jianhang Zhang, Ruibin Zhong, Bir Bhanu, Yuling Chen, Qingfeng Zhang, Haoqing Tang
2021 Sensors  
In this paper, a transmission-guided lightweight neural network called TGL-Net is proposed for efficient image dehazing.  ...  A double-error loss function that combines the errors of a transmission map with the errors of a dehazed image is used to guide network training.  ...  Data Availability Statement: The proposed TGL-Net and some testing images can be downloaded from the link https://github.com/lizhangray/TGL-Net.  ... 
doi:10.3390/s21030960 pmid:33535456 pmcid:PMC7867112 fatcat:ghiyzev5yfdrtihvczboys2c5i

Deep-Energy: Unsupervised Training of Deep Neural Networks [article]

Alona Golts and Daniel Freedman and Michael Elad
2019 arXiv   pre-print
"Deep Energy" is demonstrated in this paper on three different tasks -- seeded segmentation, image matting and single image dehazing -- exposing its generality and wide applicability.  ...  The proposed approach, termed "Deep Energy", trains a Deep Neural Network (DNN) to approximate this minimization for any chosen input.  ...  Finally, in single image dehazing the network consists of 6 dilated residual blocks with zero padding.  ... 
arXiv:1805.12355v2 fatcat:urfhf6bjwzbmjmfwlpkm42wjzy

Exploiting Raw Images for Real-Scene Super-Resolution [article]

Xiangyu Xu, Yongrui Ma, Wenxiu Sun, Ming-Hsuan Yang
2021 arXiv   pre-print
For the second issue, we develop a two-branch convolutional neural network to exploit the radiance information originally-recorded in raw images.  ...  In this paper, we study the problem of real-scene single image super-resolution to bridge the gap between synthetic data and real captured images.  ...  "sFF" and "FF" respectively represent the single-scale and multi-scale feature fusion.  ... 
arXiv:2102.01579v1 fatcat:tzanyrixcjb6dahh4dxdovancu

A review of Generative Adversarial Networks (GANs) and its applications in a wide variety of disciplines – From Medical to Remote Sensing [article]

Ankan Dash, Junyi Ye, Guiling Wang
2021 arXiv   pre-print
GANs combine two neural networks that compete against one another using zero-sum game theory, allowing them to create much crisper and discrete outputs.  ...  GANs can also be utilised for a variety of science-related activities, including protein engineering, astronomical data processing, remote sensing image dehazing, and crystal structure synthesis.  ...  Multi-Scale SSIM or MS-SSIM [174] is a multiscale version of SSIM that allows for more flexibility in incorporating image resolution and viewing conditions than a single scale approach.  ... 
arXiv:2110.01442v1 fatcat:mqpnqw2ysfdz7dneajiw33dbga

Learning Neural Textual Representations for Citation Recommendation

Binh Thanh Kieu, Inigo Jauregi Unanue, Son Bao Pham, Hieu Xuan Phan, Massimo Piccardi
2021 2020 25th International Conference on Pattern Recognition (ICPR)  
Narasimha 2233 Region and Relations Based Multi Attention Network for Graph Classification DAY 1 -Jan 12, 2021 Zhou, Shibo; Li, Xiaohua 2242 Spiking Neural Networks with Single-Spike Temporal-Coded  ...  Monocular Dense Depth Estimation with Morphology DAY 3 -Jan 14, 2021 Wei, Pan; Wang, Xin; Wang, Lei; Xiang, Ji 909 SIDGAN: Single Image Dehazing without Paired Supervision DAY 3 -Jan 14, 2021  ... 
doi:10.1109/icpr48806.2021.9412725 fatcat:3vge2tpd2zf7jcv5btcixnaikm

Deep Texture and Structure Aware Filtering Network for Image Smoothing [article]

Kaiyue Lu, Shaodi You, Nick Barnes
2018 arXiv   pre-print
To this end, we generate a large dataset by blending natural textures with clean structure-only images, and then build a texture prediction network (TPN) that predicts the location and magnitude of textures  ...  We then combine the TPN with a semantic structure prediction network (SPN) so that the final texture and structure aware filtering network (TSAFN) is able to identify the textures to remove ("texture-awareness  ...  .: Deep multi-scale convolutional neural network for dynamic scene deblurring. In: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR). (July 2017) 44.  ... 
arXiv:1712.02893v2 fatcat:kurro6nstrespny4kqskqcutua

Synthesis of High-Quality Visible Faces from Polarimetric Thermal Faces using Generative Adversarial Networks [article]

He Zhang, Benjamin S. Riggan, Shuowen Hu, Nathaniel J. Short, Vishal M.Patel
2018 arXiv   pre-print
The proposed network consists of a generator sub-network, constructed using an encoder-decoder network based on dense residual blocks, and a multi-scale discriminator sub-network.  ...  We propose a Generative Adversarial Networks (GAN) based multi-stream feature-level fusion technique to synthesize high-quality visible images from prolarimetric thermal images.  ...  . (3) Single-stream dense residual decoder. (4) Multi-scale discriminator.  ... 
arXiv:1812.05155v1 fatcat:yqhzbppyvnedviov7mtmgwjuma
« Previous Showing results 1 — 15 out of 30 results