# First, HOGSecond, star-structured deformable part: latent SVM 来学习，也就是说我们有了一张正样本的图片，标注了bounding box，我们要在某一位置，某一尺度，提取出一个区域作为某一component 的正样本。 自行车的车轮的位 ...

R-CNNPre-knowledge1. Selective SearchGoal:Propose possible regions of the object. Considerations: 尽可能遍历所有的尺度，但是不同于暴力穷举，可以先得到小尺度的区域，然后一次次合并得到大的尺寸，这样也 ...

Lifelong Machine LearningIntroductionDefinition Key Characteristics Continuous learning process Knowledge accumulation in KB Use of past knowledge to ...

A. How to install Vatic in one line codeVatic is a very convinient online annotation tool for labeling objects of videos. The vatic annotation tool en ...

PreinstallInstall CUDAInstall cuda by run fileRestart the computer! Install by deb file Install cuDNNStep 1. Download cuDNN from Nvidia cuDNN Step 2 ...

This article gives a high level overview for the process of recognizing the object by neural network. Step 1To be more clever, we could run this algor ...

In this article, I will introduce MNIST data set and review the process of training the MINIST data set to get the model by using TensorFlow. MNIST Da ...

Part I Computer Vision Field Leaders：http://mp.weixin.qq.com/s?__biz=MzAwMTA3MzM4Nw==&mid=2649439592&idx=1&sn=fdb687300e4a930fdd08c23d8816 ...

Part I: Basic Method “Histograms of Oriented Gradients for Human Detection,” N. Dalal and W. Triggs, Proc. IEEE CVPR 2005. Improvement: Detect the bo ...

I. The history of recognition II. The process of object recognition http://vision.stanford.edu/teaching/cs131_fall1617/lectures/lecture17_intro_objre ...