hadxu / machine-learning Goto Github PK

my machine-learning tutorial

Python 0.95% Jupyter Notebook 99.03% HTML 0.02% JavaScript 0.01%

machine-learning's Introduction

Hi Zhenlei 👋

Zhenlei was born in gaoyou(高邮), Jiangsu(江苏),China. He is currently working in Tencent AI Lab, and graduated in Jiangnan University in 2020.

Zhenlei research interests mainly focus on machine learning,deep learning, and distributed computing programming. He is currently learning Golang, Rust and some interesting algorithm and AI Compiler.

My Social Media 😺

website: hadxu.github.io
Github: hadxu
Linkedln: @hadxu
Twitter: @hadxu
zhihu: @hadxu

struct lei;

impl lei {
    const NAME: &'static str = "lei";
    const LOCATION: &'static str = "Shenzhen, China";
    const PROFILE: &'static str = "Algorithm Developer";
}

struct Skills;

impl Skills {
    const LANGUAGES: [&'static str; 5] = ["Python", "Rust", "C/C++", "Java", "swift", "shell"];
    const OPERATION_SYSTEMS: [&'static str; 3] = ["macOS", "Linux", "Windows"];
    const STORAGES: [&'static str; 3] = ["MongoDB", "Redis", "MySQL"];
    const Framework: [&'static str; 2] = ["Tensorflow", "Pytorch"];
    const WEB_FRAMEWORKS: [&'static str; 2] = ["Flask", "Rocket"];
    const DEVOPS: [&'static str; 2] = ["Docker", "Raspberry"];
}

machine-learning's People

Contributors

Stargazers

Watchers

Forkers

lkaige ouya-bytes eternonq joebling zhipengliu6 rhythm92 yongliangliu boozyguo lxj0276 baconsun beckett1124 iflier solosuns wangjnethca hzy-zg perfect28 hundred06 chankeh ln512 likefool allen-liang likeyu21 neeraj12121 xiangqian0011 wallace00v hjc5345 dafeix dvbug morindaz gisdeveloper2017 carabob zxkplus clustersdata qingfengting2017 zgdrzgdr pzhao16me yassinia shifengjuan1 keyman9848 attentionpassion bigrlab chenjordan leidaguo ninachang1107 eight-corner tpp000 tbaofang sunaque wanglee513 chenyyx roc-j jiangzhonglian masdude hlld3344 darcy0511 xbeeprince harryheone1 soccermouse gg00 jiangfuqiang unversityofjiangsu alan-han wchzh binkes windrise zhouzhibiao fendouai njustzsm eelva wustnlp zfxu lemonrong muhongfan gbusr kirlinl meihanwu superkey79 csfeinixwang wenyali xxxhycl2010 xiaye1254 lwkeya yspdsh mingjunw caoliangjie seebrand pete43 kyridiculous1993 lightyear0117 2013zhouying aomeyao maozhitao liuchenbaidu wangsenhong hordaway thefatty0418 xiaoluenbi ahaoboy qifu99 yunhai0920

machine-learning's Issues

Python files

Is this based on tensorflow? I didnot find the py file in cloning files.

可否应用于单通道

您好，如果我有的是灰度图，是否只需要修改input_shape里的通道值为1就行了吗，预训练的权重还能使用吗

代码疑问

def load_image(imageurl):
    im = cv2.resize(cv2.imread(imageurl),(224,224)).astype(np.float32)
    im[:,:,0] -= 103.939
    im[:,:,1] -= 116.779
    im[:,:,2] -= 123.68
    im = im.transpose((2,0,1))
    im = np.expand_dims(im,axis=0)
    return I'm

请问这个为什么三个通道要减去这个值？
后面的转置是什么目的？不太懂，希望大神解答下谢谢！

正确率吓人

加载的vgg模型来预测的时候都不准确，不知道是怎么回事
example：predict('zebra.jpg')
result:n12144580 corn

            predict('pig.jpg')
            result:n02325366 wood rabbit, cottontail, cottontail rabbit

face-emotion

请问你是在什么环境下跑的？人脸情感识别是在tensorflow+keras下么？还是theno+keras，我在tensorflow跑这个程序，提示加载优化器错误

vgg16_weights.h5加载权值出错

我用的是windows7+keras+tensorflow,但是无法使用您的vgg16_weight.h5，请问您使用的是Th后端训练的vgg16吗？一下子就掉坑里，贼爽

关于dive_into_keras_vgg16的预测脚本问题

你好，目录下的ex.py文件可以跑过，但是后面预测的那部分源码在哪里，感觉看的不是很懂。能否放上完整的脚本供参考，非常感谢。

现在我们开始来预测了

首先写一个方法来加载并处理图片

def load_image(imageurl):
    im = cv2.resize(cv2.imread(imageurl),(224,224)).astype(np.float32)
    im[:,:,0] -= 103.939
    im[:,:,1] -= 116.779
    im[:,:,2] -= 123.68
    im = im.transpose((2,0,1))
    im = np.expand_dims(im,axis=0)
    return im

读取vgg16的类别文件

f = open('synset_words.txt','r')
lines = f.readlines()
f.close()

def predict(url):
    im = load_image(url)
    pre = np.argmax(model.predict(im))
    print lines[pre]

%pylab inline

Populating the interactive namespace from numpy and matplotlib

from IPython.display import Image

Image('cat1.jpg')