- 博客(29)
- 收藏
- 关注
原创 PCA主成分分析去噪与降维
一、目的1、降低维数(主要目的)2、去噪二、引用https://www.cnblogs.com/sweetyu/p/5085798.html
2020-06-05 10:44:57 19928 1
原创 f-GAN: Training Generative Neural Samplers using Variational Divergence Minimization
fff-GAN: Training Generative Neural Samplers using Variational Divergence MinimizationPaper:http://papers.nips.cc/paper/6066-f-gan-training-generative-neural-samplers-using-variational-divergence-min...
2020-05-27 16:11:05 751
原创 InfoGAN:Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial NetsPaper:https://arxiv.org/abs/1606.03657Code:https://github.com/openai/InfoGANTips:Nips2016的一篇paper...
2020-05-27 16:11:00 532
原创 WGAN and WGAN-GP:Wasserstein GAN and Improved Training of Wasserstein GANs
Wasserstein GANPaper:https://arxiv.org/pdf/1701.07875.pdfCode:https://github.com/igul222/improved_wgan_training参考:https://lilianweng.github.io/lil-log/2017/08/20/from-GAN-to-WGAN.htmlhttps://vinc...
2020-05-27 16:10:54 472
原创 DeblurGAN:Blind Motion Deblurring Using Conditional Adversarial Networks
DeblurGAN:Blind Motion Deblurring Using Conditional Adversarial NetworksPaper:http://openaccess.thecvf.com/content_cvpr_2018/papers/Kupyn_DeblurGAN_Blind_Motion_CVPR_2018_paper.pdfCode:https://githu...
2020-05-27 16:10:46 1198 1
原创 patch-GAN(pixel2pixel):Image-to-Image Translation with Conditional Adversarial Networks
Image-to-Image Translation with Conditional Adversarial NetworksPaper:https://arxiv.org/pdf/1611.07004.pdfCode:https://github.com/affinelayer/Pix2Pix-tensorflowTips:CVPR2017的一篇paper。(阅读笔记)...
2020-05-27 16:10:21 1196
原创 Perceptual Losses for Real-Time Style Transfer and Super-Resolution
Perceptual Losses for Real-Time Style Transfer and Super-ResolutionPaper:https://arxiv.org/pdf/1603.08155.pdfCode:https://github.com/ChengBinJin/Real-time-style-transferSupplement:https://cs.stanfo...
2020-05-27 16:09:58 365
原创 Pre-training on Grayscale ImageNet Improves Medical Image Classification
Pre-training on Grayscale ImageNet Improves Medical Image ClassificationPaper:http://openaccess.thecvf.com/content_ECCVW_2018/papers/11134/Xie_Pre-training_on_Grayscale_ImageNet_Improves_Medical_Imag...
2020-03-16 21:08:58 273
原创 MINE:Mutual Information Neural Estimation
Mutual Information Neural EstimationPaper:https://arxiv.org/pdf/1801.04062v4.pdfCode:https://github.com/mzgubic/MINETips:ICML2018的一篇paper。(阅读笔记)1.Main idea高维连续随机变量的互信息的估计可以被神经网络通过梯度下降实现。the est...
2020-02-28 22:40:50 2402 2
原创 DCGAN:Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
Unsupervised Representation Learning with Deep Convolutional Generative Adversarial NetworksPaper:https://arxiv.org/pdf/1511.06434v2.pdfCode:https://github.com/znxlwm/tensorflow-MNIST-GAN-DCGANTips...
2020-02-24 22:18:34 635
原创 Reptile: On First-Order Meta-Learning Algorithms
On First-Order Meta-Learning AlgorithmsPaper:https://arxiv.org/pdf/1803.02999.pdfCode:https://github.com/openai/supervised-reptileTips:OpenAi的一篇相似MAML的Meta-learning相关的paper。(阅读笔记)1.Main idea目标旨...
2020-02-18 20:34:26 916
原创 MAML:Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
Model-Agnostic Meta-Learning for Fast Adaptation of Deep NetworksPaper:https://arxiv.org/pdf/1703.03400.pdfCode:https://github.com/cbfinn/mamlTips:ICML的一篇Meta-learning相关的paper,可参考李宏毅的公开课。(阅读笔记)1....
2020-02-15 23:52:10 407
原创 K-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation
K-SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse RepresentationPaper:http://sites.fas.harvard.edu/~cs278/papers/ksvd.pdf参考周志华《机器学习》(阅读笔记,如有不当,敬请谅解)1、写在前面该篇论文是一篇methodology的论...
2020-01-19 18:37:42 1492
原创 范数的数学意义
L0,L1,L2范数的数学意义(如有不当,敬请斧正)Tips:范数所表示的一些数学意义:众数,中位数,均值A:\mathcal{A:}A: L0范数:求L0范数最小时,表示的是数据中的众数modes(假设00=00^0=000=0的条件下)。其中Y:{y1,y2,...,yk}\mathcal{Y}:\{{y_1,y_2,...,y_k}\}Y:{y1,y2,...,yk}是数据...
2020-01-15 17:31:38 903
原创 Toward Convolutional Blind Denoising of Real Photographs
Toward Convolutional Blind Denoising of Real PhotographsPaper:http://openaccess.thecvf.com/content_CVPR_2019/papers/Guo_Toward_Convolutional_Blind_Denoising_of_Real_Photographs_CVPR_2019_paper.pdfCo...
2020-01-09 14:44:22 401
原创 Pytorch之卷积神经网络(Mnist)学习
机器学习的Github代码阅读总结以前写的代码太差,而代码的书写和可读性很重要,在此学习记录。阅读的代码是pytorch官方给的例子:https://github.com/pytorch/examples1、Argparse库的使用Argparse库的使用很频繁,一般是写在开头,主要是用于命令行(cmd)中运行python文件的一些参数设置(如:batch size, learning ra...
2019-09-20 15:43:30 662
原创 Instance Normalization: The Missing Ingredient for Fast Stylization
Instance Normalization: The Missing Ingredient for Fast StylizationPaper:https://arxiv.org/pdf/1607.08022v3.pdfhttps://arxiv.org/pdf/1701.02096.pdfCode:https://github.com/lengstrom/fast-style-tran...
2019-09-03 23:55:39 1299
原创 U-Net: Convolutional Networks for Biomedical Image Segmentation
U-Net: Convolutional Networks for Biomedical Image SegmentationPaper:https://arxiv.org/pdf/1505.04597v1.pdf1.AimWe present a network and training strategy that relies on the strong use of data augm...
2019-09-02 11:19:52 288
原创 Noise2noise: Learning image restoration without clean data
Noise2noise: Learning image restoration without clean dataPaper:https://arxiv.org/abs/1803.04189Code:https://github.com/NVlabs/noise2noise1.Main ideaIt is possible to learn to restore images by on...
2019-08-31 22:12:19 895
原创 Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Skip Connections
Image Restoration Using Very Deep Convolutional Encoder-Decoder Networks with Symmetric Skip ConnectionsPapers:Nips:https://papers.nips.cc/paper/6172-image-restoration-using-very-deep-convolutional-...
2019-08-30 16:44:08 776
原创 matlab图像模糊和亮度不均学习记录
最近涉及到有关图像处理的知识,在此进行学习记录。本人是通过matlab进行实现的1、图像模糊matlab相关文档链接:https://ww2.mathworks.cn/help/images/ref/fspecial.html参考博客:https://blog.csdn.net/ji0525084/article/details/8061295通过matlab实现模糊主要就是两类,运动模...
2019-06-21 08:57:23 1841
原创 索引图像读取与分析理解
索引图像最近在做语义分割相关的内容,但是一直对数据集(如voc2012)的图像格式不太清楚,查阅资料在此进行记录。相关链接:1、索引图像概念:https://baike.baidu.com/item/索引图像2、博客解释:www.packjour.cn/bzgcgk/ch/index.aspx1、索引图像读取索引图像的读取是用matlab进行读取的,方式如下:(本人是加载的是VOC2...
2019-06-13 20:11:20 5756 1
原创 tf.train.shuffle_batch打乱顺序输出数据
tf.train.shuffle_batch该函数怎么输出数据的?tf.train.shuffle_batch主要有三个参数:batch_size,capacity,min_after_dequeue。这三个数据决定了怎么输出数据,数据有多乱。首先我们知道tfrecord输出是按照队列输出的,那么我们一次从tfrecord加多少容量的队列呢?这个参数就是capacity。注意,此时首先加载一...
2019-05-07 22:07:19 2154 2
原创 tf.nn.conv2d_transpose反卷积(转置卷积)
tf.nn.conv2d_transpose反卷积(转置卷积)查阅资料了许久,很多博客资料都对知识进行详细讲解,但一到strides,padding就一带而过,自己发现还是对于步长stride和填充padding这两个因素对输出张量大小的关系不是很清楚,于是自行代码实验并在此记录。反卷积的基本资料查阅:1、介绍了反卷积的各种详细的知识: https://blog.csdn.net/mao_x...
2019-05-03 23:05:20 8836 3
原创 深度学习框架Tensorflow学习
一、基本概念1、张量(Tensor)张量是tensorflow程序中的基本成分,有了张量才能进行过后的一系列的运算。但是它表示的是所有的数据的运算的过程,而并不是数据本身。如下面代码,定义了a,b两个常量的张量,并通过相加得到了result。但是此时的result结果并不是[3., 5.],而是result张量的基本信息(名字,维度,类型),我们打印result结果如下。import te...
2019-03-27 19:45:05 156
原创 STM32F4的电压实时检测
本人最近课程设计学习STM32F4系列,在此进行学习记录!本次记录完成的结课项目!课程实训涉及STM32基本知识,GPIO,外部中断EXTI,串口USART,ADC以及DMA,最后贴有注解代码!一、任务A:按键控制ADC开始工作B:通过旋转滑动变阻器改变电压,随着电压的不同,通过点亮LED灯发出警告,当电压达到最大值,蜂鸣器叫。C:串口输出当前电压值。二、过程1、代码风格考虑该项目...
2019-03-06 09:05:04 6433 1
原创 决策树和随机森林原理(Decision Trees and Random Forests)
本人最近涉及到有关决策树知识,在此进行学习记录。本人是基于周志华《机器学习》来学习的!相当于也是个读书笔记!一、基本流程一棵决策树一般包含一个根结点、若干个内部结点和若干个叶结点。叶结点对应于决策结果,其他每个结点则对应于一个属性测试。很显然的那么根结点就包含了样本全集。决策树学习的目的是为了产生一棵泛化能力强,即处理未见示例能力强的决策树。如上图算法流程所示,训练集D中每一类即对应了...
2019-01-30 10:51:39 678
原创 Windows下安装配置Tensorflow
Windows下安装配置TensorFlow个人只亲测了cpu版本,在此就分享个人安装配置cpu版本的tnsorflow的方法。使用版本 Windows7 64位一、安装AnacondaAanconda这个软件究竟怎么样,我就不怎么说了,反正就是方便!!!直接上官方网站:https://www.anaconda.com/download/直接选择Download进行下载,下载完成后...
2018-10-28 21:56:46 340
空空如也
空空如也
TA创建的收藏夹 TA关注的收藏夹
TA关注的人