http://www.indiana.edu/~p1013447/dictionary/lat_i.htm

【This means that neighboring visual neurons respond LESS if they are activated at the same time than if one is activated alone. So the fewer neighboring neurons stimulated, the more strongly a neuron responds. 该神经元周围的同时被刺激的神经元数目越少,该神经元的响应就越显著。】

Lateral inhibition refers to the inhibition that neighboring neurons in brain pathways have on each other. For example, in the visual system, neighboring pathways from the receptors to the optic nerve, which carries information to the visual areas of the brain, show lateral inhibition. This means that neighboring visual neurons respond LESS if they are activated at the same time than if one is activated alone. So the fewer neighboring neurons stimulated, the more strongly a neuron responds.

You might expect that such inhibition would decrease the visual system's ability to represent information. In fact, this process greatly increases the visual system's ability to respond to edges of a surface. This happens because neurons responding to the edge of a stimulus respond more strongly than do neurons responding to the middle. The "edge" neurons receive inhibition only from neighbors on one side -- the side away from the edge. Neurons stimulated from the middle of a surface get inhibition from all sides.

【It makes edges stand out 强化边缘】

The pictures you see taken from space craft are computer enhanced by a process just like lateral inhibition. Computer transformation of the raw image makes neighboring points of the computerized image inhibit each other. This makes the very faint edges in the original image much sharper. This is why lateral inhibition is important: It makes edges stand out

lateral inhibition的更多相关文章

  1. Paper/ Overview | CNN(未完待续)

    目录 I. 基础知识 II. 早期尝试 1. Neocognitron, 1980 2. LeCun, 1989 A. 概况 B. Feature maps & Weight sharing ...

  2. AlexNet

    AlexNet学习笔记 目录 AlexNet整体结构 CNN 全连接 TensorFlow实现 AlexNet是2012年ImageNet竞赛冠军获得者Hinton和他的学生Alex Krizhevs ...

  3. 思考卷积神经网络(CNN)中各种意义

    原文:https://blog.csdn.net/aimreant/article/details/53145063 思考卷积神经网络(CNN)中各种意义 只是知道CNN是不够,我们需要对其进行解剖, ...

  4. AlexNet论文翻译-ImageNet Classification with Deep Convolutional Neural Networks

    ImageNet Classification with Deep Convolutional Neural Networks 深度卷积神经网络的ImageNet分类 Alex Krizhevsky ...

  5. caffe中的Local Response Normalization (LRN)有什么用,和激活函数区别

    http://stats.stackexchange.com/questions/145768/importance-of-local-response-normalization-in-cnn ca ...

  6. caffe 学习(3)——Layer Catalogue

    layer是建模和计算的基本单元. caffe的目录包含各种state-of-the-art model的layers. 为了创建一个caffe model,我们需要定义模型架构在一个protocol ...

  7. Local Response Normalization 60 million parameters and 500,000 neurons

    CNN是工具,在图像识别中是发现图像中待识别对象的特征的工具,是剔除对识别结果无用信息的工具. ImageNet Classification with Deep Convolutional Neur ...

  8. Image Processing and Analysis_8_Edge Detection:Statistical edge detection_ learning and evaluating edge cues——2003

    此主要讨论图像处理与分析.虽然计算机视觉部分的有些内容比如特 征提取等也可以归结到图像分析中来,但鉴于它们与计算机视觉的紧密联系,以 及它们的出处,没有把它们纳入到图像处理与分析中来.同样,这里面也有 ...

  9. 1 - ImageNet Classification with Deep Convolutional Neural Network (阅读翻译)

    ImageNet Classification with Deep Convolutional Neural Network 利用深度卷积神经网络进行ImageNet分类 Abstract We tr ...

随机推荐

  1. 标准C程序设计七---114

    Linux应用             编程深入            语言编程 标准C程序设计七---经典C11程序设计    以下内容为阅读:    <标准C程序设计>(第7版) 作者 ...

  2. linux 时间模块 一

    作者:云天明链接:https://www.zhihu.com/question/50764115/answer/122613742来源:知乎著作权归作者所有.商业转载请联系作者获得授权,非商业转载请注 ...

  3. android 设置app root权限简单方法

    vim frameworks/base/core/java/com/android/internal/os/ZygoteConnection.java +709 private static void ...

  4. luogu P1040 加分二叉树

    题目描述 设一个n个节点的二叉树tree的中序遍历为(1,2,3,-,n),其中数字1,2,3,-,n为节点编号.每个节点都有一个分数(均为正整数),记第i个节点的分数为di,tree及它的每个子树都 ...

  5. springboot 集成 freemarker

    前面我们已经实现了thymeleaf模板,其实freemarker和thymeleaf差不多,都可以取代JSP页面,实现步骤也差不多,我们来简单实现一下 引入pom.xml依赖如下 <depen ...

  6. 小W计树

    排列组合思想. 先跑一遍最短路, 再从1节点开始搜索, 假如搜到一个点的路径长度等于最短路, 则记录到达该点的路径数 + 1. 最后遍历一遍, ans *= rec[i] 输出答案即可. 关键在于想到 ...

  7. 混合背包 hdu5410 CRB and His Birthday

    传送门:点击打开链接 题意:你有M块钱,如今有N件商品 第i件商品要Wi块,假设你购买x个这种商品.你将得到Ai*x+Bi个糖果 问能得到的最多的糖果数 思路:很好的一道01背包和全然背包结合的题目 ...

  8. RTMP协议规范(转载)

    译序: 本文是为截至发稿时止最新 Adobe 官方公布的 RTMP 规范.本文包含 RTMP 规范的全部内容.是第一个比较全面的 RTMP 规范的中译本.由于成文时间仓促,加上作者知识面所限,翻译错误 ...

  9. Node.js知识点学习

    Node.js知识点学习 一.基本概念 Node.js,或者 Node,是一个可以让 JavaScript 运行在服务器端的平台.可以说,Node.js开创了javascript模块化开发的先河,早期 ...

  10. Ejb in action(六)——拦截器

    Ejb拦截器可以监听程序中的一个或全部方法.与Struts2中拦截器同名,并且他们都可以实现切面式服务.同一时候也与Spring中的AOP技术类似. 不同的是struts2的拦截器的实现原理是一层一层 ...