▍1. DenseNet201
说明: DenseNet201卷积神经网络训练模型,python语言编程,粗略训练,入门难度。(DenseNet201 convolutional neural network training model, python programming, rough training, entry difficulty.)
说明: DenseNet201卷积神经网络训练模型,python语言编程,粗略训练,入门难度。(DenseNet201 convolutional neural network training model, python programming, rough training, entry difficulty.)
说明: shape_predictor_68_face_landmarks.dat文件太大,请自行下载(face recognition,python,ssh)
说明: 卷积神经网络原理介绍,实验使用,仅供参考(The principle of convolution neural network is introduced and used in experiment for reference only)
说明: 无监督训练的生成对抗网络程序,python版本,生成有关光照不同的图片(Unsupervised training generated against the network program, python version, to generate different pictures about the light)
说明: Delm,可实现分类的功能,具有训练速度快,使用简单等优点。(Elm_kernel, which can realize classification)
说明: 利用该编码对声源位置进行了“端到端的深度学习:从音频信号到声源位置坐标”的定位(By using this code, the "end-to-end deep learning: from audio signal to sound source location coordinate" is carried out)
说明: 基于DOA的水声网络目标定位方法,首先利用MUSIC算法进行测向,再利用最小二乘法进行融合以定位。(DOA based underwater acoustic network target location method, first use music algorithm for direction finding, then use least square method for fusion to locate.)
说明: 这是一个抢票的python程序,效果非常好(This is a python program for ticket grabbing. The effect is very good)
说明: ABAQUS喷蚀模拟,对实际的丸粒冲蚀过程进行模拟,效率高,可靠性好(ABAQUS erosion simulation)
说明: 基于勾勒图的图像检索源码(download from github)(code for DeepSBIR-master which is download from github)
说明: 猫狗大战的代码,图片分类,二分类,图片网上下载即可(cat_vs_dog,picture classification, pictures can be downloaded online)
说明: 蒙特卡洛编程,含程序代码,代码直接复制即可,还有实例!!!(Monte Carlo programming, including program code.)
说明: 一个多核学习python版本的代码,里面有平均核方法(A python library of multiple kernel learning)
以图像处理和knn算法为基础,利用knn算法,在python开发环境中设计了一个手写数字识别系统,该系统采用28*28的模板对图新分成784块提取手写数字的784个特征,在二值图像数据基础上,通过朴素贝叶斯分类器算法,对0~9这十类手写数字进行学习和分类。(Based on image processing and KNN algorithm, a handwritten digit recognition system is designed in Python development environment by using KNN algorithm. The system uses 28*28 template to divide the graph into 784 blocks to extract 784 features of handwritten digits. On the basis of binary image data, it learns and classifies the ten types of handwritten digits, 0-9, through Naive Bayesian classifier algorithm.)