GithubHelp home page GithubHelp logo

bubbliiiing / keras-face-recognition Goto Github PK

View Code? Open in Web Editor NEW
221.0 8.0 56.0 2.45 MB

这是一个基于mtcnn和facenet的人脸识别模型,可实现在线人脸识别。

License: MIT License

Python 100.00%

keras-face-recognition's Introduction

Face-Recognition:人脸识别算法在Keras当中的实现


目录

  1. 所需环境 Environment
  2. 文件下载 Download
  3. 使用方法 Usage
  4. 效果 Performance

所需环境

tensorflow-gpu==1.13.1
keras==2.1.5

文件下载

进行预测所需的facenet_keras.h5可以在Release里面下载。
也可以去百度网盘下载
链接: https://pan.baidu.com/s/1A9jCJa_sQ4D3ejelgXX2RQ 提取码: tkhg

使用方法

1、先将整个仓库download下来。
2、下载完之后解压,同时下载facenet_keras.h5文件。
3、将facenet_keras.h5放入model_data中。
4、将自己想要识别的人脸放入到face_dataset中。
5、运行face_recognize.py即可。
6、align.py可以查看人脸对齐的效果。

效果

face_recognize.py的运行结果:
result)

keras-face-recognition's People

Contributors

bubbliiiing avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

keras-face-recognition's Issues

tensorflow-gpu版本

tensorflow-gpu这个版本太低了吧,和我的cuda不匹配,重新安装cuda好麻烦,有没有别的办法啊,或者有没有其他可以实时人脸识别的代码模型了呀,救救孩子吧

貌似环境兼容问题

原始报错TypeError: Cannot interpret feed_dict key as Tensor: Tensor Tensor("Placeholder:0", shape=(3, 3, 3, 10), dtype=float32) is not an element of this graph.

加了```
from keras import backend as K
K.clear_session()

后报错:
`ValueError: Tensor("Placeholder:0", shape=(3, 3, 3, 10), dtype=float32) must be from the same graph as Tensor("conv1/kernel:0", shape=(3, 3, 3, 10), dtype=float32_ref).`

运行中出现的问题

运行face_recognize.py时79行:height,width,_ = np.shape(draw) 140行:dududu.recognize(draw)
发生异常: ValueError
not enough values to unpack (expected 3, got 0)
请问怎么解决?

模型训练

你好博主,请问想用其他网络替换其中的inception_resenet 模型,该如何训练模型?有训练模型额代码参考么?

使用中一些问题的解决

以下仅个人拙见,也许有更好的解决方案:

  1. 在安装所有模块后仍然报错(具体是啥我忘了),反正是一个decode错误,解决方法是将h5py降级到 2.10.0 。
  2. 人脸识别的错误率挺高的,经常把我和我同学搞混😭,解决方案是在人脸库里加入一些未知人脸(即Unknown),后面可以跟一串数字,如果希望在显示中只显示Unknown而不显示数字,可以用x.isdigit()函数进行滤除。
  3. 暗光条件下表现不太好,解决方案是使用对比度调整,对比测试后发现伽马变换效果最好,伽马指数在0.75~0.8时比较稳定。
  4. 因为我要做一个签到系统,所以为了减少签到错误,我加了一个验证,连续5~10次出现同一个识别结果时才在画面中显示结果,这样虽然有延迟,但是识别准确率提高了不少。

数据库图片训练中的一个问题

当我加入多张图片进入数据库时,出现了这个错误:

Traceback (most recent call last):
File "G:/Python_Project/Tensorflow_FaceNet_one/face_recognize.py", line 136, in
dududu = face_rec()
File "G:/Python_Project/Tensorflow_FaceNet_one/face_recognize.py", line 56, in init
crop_img, _ = utils.Alignment_1(crop_img,landmark)
File "G:\Python_Project\Tensorflow_FaceNet_one\utils\utils.py", line 232, in Alignment_1
new_img = cv2.warpAffine(img, RotationMatrix, (img.shape[1], img.shape[0]))
cv2.error: OpenCV(4.5.1) C:\Users\appveyor\AppData\Local\Temp\1\pip-req-build-i1s8y2i1\opencv\modules\imgproc\src\imgwarp.cpp:2595: error: (-215:Assertion failed) src.cols > 0 && src.rows > 0 in function 'cv::warpAffine'


该错误发生在openCV的warpAffine的翻转(仿射转换)中,我查阅了许多资料,绝大多数结论认为该错误的出现是源于序列错误或则路径错误,我找不到可以修复它的方式,我对此感到迷惑,希望能得到您的回复。

代码不全啊

请问您的训练代码怎么没有呢,我在bilibili上看了您的讲解,但这里没有您构建模型,以及训练的源码啊,我想使用您的源码做一些改动以便适应我现在的研究。可以分享完整的代码出来吗,感谢呢

训练权重

您好,我发现model_data文件夹下面只有p r o三个网络的权重,没有facenet的网络权重,您能重新放一下facenet网络的权重文件吗?谢谢大佬!

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.