![](https://img-blog.csdnimg.cn/b79e0a8b19314fdd8097a0c765bb2542.png?x-oss-process=image/resize,m_fixed,h_224,w_224)
图像处理
文章平均质量分 87
随便写写就完了
St-sun
Because it's everything
展开
-
LIME Low light Image Enhancement via Illumination Map Estimation
通过寻找RGB三通道中的最大值,并分别估计每个像素的illumination,然后添加一个结构来完善initialilluminationmap来作为最终的illuminationmap。该方法具有LIME的有效性,并具有很好的图像增强效果和效率。高可见度的图像具有目标场景的具体细节,low-light情况下的图像可见度低,基于Retinex策略估计illuminationmap来增强低光照。本方法只估计illumination缩小solutionspace并减少计算成本。...原创 2022-07-27 13:57:20 · 2390 阅读 · 0 评论 -
python色彩变换CMYK,RGB,HSI
python色彩变换CMYK,RGB,HSI原创 2022-07-25 17:15:23 · 2753 阅读 · 0 评论 -
PR-RL:Portrait Relighting via Deep Reinforcement Learning
PR-RL: Portrait Relighting via Deep Reinforcement Learning原创 2022-05-27 13:06:24 · 555 阅读 · 2 评论 -
ReLLIE Deep Reinforcement Learning for Customized Low-Light Image Enhancement
文章目录ReLLIE Deep Reinforcement Learning for Customized Low-Light Image Enhancement0 Abstract and Introduction2 Approach2.1 LLIE via Curve Adjustment2.2 LLIE as Markov Decision Process3 Proposed RELLIE3.1 Agent3.2 Non-Reference Losses3.3 Channel Dependent M原创 2022-04-17 22:29:48 · 2052 阅读 · 0 评论 -
sobel_GxGy边缘图_方向梯度直方图
Sobel滤波边缘提取1、设计能够提取X方向、Y方向的Sobel滤波器,将上述滤波器和图像进行卷积,生成X方向和Y方向边缘图像。思想:X方向卷积因子:[−10+1−20+2−10+1]\left[ \begin{matrix}-1&0&+1\\ -2&0&+2\\-1&0&+1 \end{matrix} \right]⎣⎡−1−2−1000+1+2+1⎦⎤Y方向卷积因子:[+1+2+1000−1−2−1]\left[ \begin{matri原创 2022-04-17 11:53:56 · 365 阅读 · 0 评论 -
空间滤波-随机椒盐噪声-高斯噪声-均值滤波器-中值滤波器
空间滤波-随机椒盐噪声-高斯噪声-均值滤波器-中值滤波器原创 2022-04-09 16:12:58 · 736 阅读 · 0 评论 -
精读A Hierarchical Reinforced Sequence Operation Method for Unsupervised Text Style Transfer
在HRL(强化学习)框架中,提出一种基于序列操作PTO(Point-Then-Operate):高级agent提出操作位置,低级agent修改句子。用于无监督文本样式的传输。 转换过程被建模为对输入句子的明确修改,这提高了可解释性、内容保存和可控的风格与内容的权衡。基于策略的训练算法来建立文本风格转换的fluency(a language model reward)、style polarity(a classification confidence reward和auxiliary classifica原创 2022-04-05 20:43:25 · 1113 阅读 · 0 评论 -
A Hierarchical Reinforced Sequence Operation Method for Unsupervised Text Style Transfer
文章目录A Hierarchical Reinforced Sequence Operation Method for Unsupervised Text Style Transfer(一种无监督文本样式转换的分层增强序列操作方法 )0 Abstract:1 Introduction2 Related WorkText Style TransferSequence Operation MethodsHierarchical Reinforcement Learning3 Formulation4 Our原创 2022-03-27 21:10:32 · 1318 阅读 · 1 评论 -
Coarse-to-Fine Region Selection and Matching
文章目录Coarse-to-Fine Region Selection and Matching0 Abstract1 Contribution2 Coarse-to-Fine Matching Algorithm2.1 Coarse-to-Fine Region Selection2.2 Coarse-to-Fine Affine Search2.3 Cost Savings of the Affine Hierarchy3 Theoretical Justification3.1 Justificati原创 2022-03-24 19:38:03 · 4496 阅读 · 0 评论 -
Park_Distort-and-Recover_Color_Enhancement_CVPR_2018_paper
文章目录Distort-and-Recover: Color Enhancement using Deep Reinforcement Learning(失真与恢复:使用深度强化学习的颜色增强 )0 Abstract1 Contribution2 Problem Formulation3 Automatic Color Enhancement3.1 Actions3.2 Features3.3 Agent4 ‘Distort-and-Recover’ Training Scheme5 Experiment原创 2022-03-15 23:42:44 · 3599 阅读 · 0 评论 -
Fully Convolutional Network with Multi-Step Reinforcement Learning
文章目录Fully Convolutional Network with Multi-Step Reinforcement Learningfor Image Processing(基于多步强化学习的全卷积网络图像处理 )0 Abstract1 Contribution2 Background3 Reinforcement Learning with Pixel-wise Rewards (PixelRL)(用像素激励强化学习)4 Reward Map Convolution (激励映射卷积)5 Appl原创 2022-03-13 18:21:39 · 4075 阅读 · 0 评论 -
HLA-Face Joint High-Low Adaptation for Low Light Face Detection
文章目录HLA-Face: Joint High-Low Adaptation for Low Light Face Detection(HLA-Face联合高低自适应微光人脸检测)0 Abstract1 Contribution2 Joint Adaptation for Dark Face Detection2.1 Motivation2.2 Bidirectional Low-Level AdaptationBrighteningNoise SynthesisColor Jittering2.3 Mu原创 2022-03-11 18:12:10 · 4422 阅读 · 0 评论 -
Moran_DeepLPF_Deep_Local_Parametric_Filters_for_Image_Enhancement_CVPR_2020_paper
文章目录DeepLPF: Deep Local Parametric Filters for Image Enhancement0 Abstract1 Contribution2 Deep Local Parameter Filters(DeepLPF)2.1 DeepLPF Architecture2.2 Local Parameter Filters2.2.1Filter Parameter Prediction Network2.2.2 Graduated Filter2.2.3 Elliptic原创 2022-03-09 20:57:07 · 2060 阅读 · 0 评论 -
Xu_Learning_to_Restore_Low-Light_Images_via_Decomposition-and-Enhancement_CVPR_2020_paper
文章目录Learning to Restore Low-Light Images via Decomposition-and-Enhancement(学习通过分解和增强恢复弱光图像)0 Abstract1 Contribution2 Proposed Model2.1 ACE Module2.2 CDT Module2.3 Proposed Dataset2.4 Training3 ConclusionLearning to Restore Low-Light Images via Decompositi原创 2022-03-07 20:53:45 · 1346 阅读 · 0 评论 -
Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement
文章目录Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement(微光图像增强的零参考深曲线估计 )BackgroundAbstractContributionLight-Enhancement Curve(LE-curve)Higher-Order Curve 高阶光照增强曲线Pixel-Wise Curve 像素级图像曲线DCE-NetNon-Reference Loss Functions 非引用损失函数Spatial原创 2022-03-05 16:34:53 · 1967 阅读 · 0 评论