FACENET

     

A TensorFlow backed FaceNet implementation for Node.js, which can solve face _verification_, _recognition_ and _clustering_ problems.

!Google Facenet

FaceNet is a deep convolutional network designed by Google, trained to solve face verification, recognition and clustering problem with efficiently at scale.

1. directly learns a mapping from face images to a compact Euclidean space where distances directly correspond to a measure of face similarity.
1. optimize the embedding face recognition performance using only 128-bytes per face.
1. achieves accuracy of 99.63% on Labeled Faces in the Wild (LFW) dataset, and 95.12% on YouTube Faces DB.

INSTALL

``shell
$ npm install facenet numjs flash-store
`

$3


1. numjs
1. flash-store

EXAMPLE

The follow examples will give you some intuitions for using the code.

1. demo exmaple will show you how to do align for face alignment and embedding to get face feature vector.
1. visualize example will calculate the similarity between faces and draw them on the photo.

1. Demo for API Usage

TL;DR: Talk is cheap, show me the code!

`ts
import { Facenet } from 'facenet'

const facenet = new Facenet()

// Do Face Alignment, return faces
const imageFile = ${__dirname}/../tests/fixtures/two-faces.jpg
const faceList = await facenet.align(imageFile)

for (const face of faceList) {
console.log('bounding box:', face.boundingBox)
console.log('landmarks:', face.facialLandmark)

// Calculate Face Embedding, return feature vector
const embedding = await facenet.embedding(face)
console.log('embedding:', embedding)
}
faceList[0].embedding = await facenet.embedding(faceList[0])
faceList[1].embedding = await facenet.embedding(faceList[1])
console.log('distance between the different face: ', faceList[0].distance(faceList[1]))
console.log('distance between the same face: ', faceList[0].distance(faceList[0]))
`

Full source code can be found at here:

The output should be something like:

`shell
image file: /home/zixia/git/facenet/examples/../tests/fixtures/two-faces.jpg
face file: 1-1.jpg
bounding box: {
p1: { x: 360, y: 95 },
p2: { x: 589, y: 324 }
}
landmarks: {
leftEye: { x: 441, y: 181 },
rightEye: { x: 515, y: 208 },
nose: { x: 459, y: 239 },
leftMouthCorner: { x: 417, y: 262 },
rightMouthCorner: { x: 482, y: 285 }
}
embedding: array([ 0.02453, 0.03973, 0.05397, ..., 0.10603, 0.15305,-0.07288])

face file: 1-2.jpg
bounding box: {
p1: { x: 142, y: 87 },
p2: { x: 395, y: 340 }
}
landmarks: {
leftEye: { x: 230, y: 186 },
rightEye: { x: 316, y: 197 },
nose: { x: 269, y: 257 },
leftMouthCorner: { x: 223, y: 273 },
rightMouthCorner: { x: 303, y: 281 }
}
embedding: array([ 0.03241, -0.0737, 0.0475, ..., 0.07235, 0.12581,-0.00817])
`

2. Visualize for Intuition

!FaceNet Visualization

1. Face is in the green rectangle.
1. Similarity(distance) between faces showed as a number in the middle of the line.
1. To identify if two faces belong to the same person, we could use an experiential threshold of distance: 0.75.

`shell
$ git clone git@github.com:zixia/node-facenet.git
$ cd facenet
$ npm install
$ npm run example:visualize

01:15:43 INFO CLI Visualized image saved to: facenet-visulized.jpg
`

3. Get the diffence of two face

Get the two face's distance, the smaller the number is, the similar of the two face

`ts
import { Facenet } from 'facenet'

const facenet = new Facenet()
const imageFile = ${__dirname}/../tests/fixtures/two-faces.jpg

const faceList = await facenet.align(imageFile)
faceList[0].embedding = await facenet.embedding(faceList[0])
faceList[1].embedding = await facenet.embedding(faceList[1])
console.log('distance between the different face: ', faceList[0].distance(faceList[1]))
console.log('distance between the same face: ', faceList[0].distance(faceList[0]))
`
Output:
distance between the different face: 1.2971515811057608
distance between the same face: 0

In the example,
faceList[0] is totally the same with faceList[0], so the number is 0
faceList[1] is different with faceList[1], so the number is big.
If the number is smaller than 0.75, maybe they are the same person.

Full source code can be found at here:

4. Save the face picture from a picture

Recognize the face and save the face to local file.

`ts
import { Facenet } from 'facenet'

const facenet = new Facenet()
const imageFile = ${__dirname}/../tests/fixtures/two-faces.jpg

const faceList = await facenet.align(imageFile)
for (const face of faceList) {
await face.save(face.md5 + '.jpg')
console.log(save face ${face.md5} successfuly)
}
console.log(Save ${faceList.length} faces from the imageFile)
`

Full source code can be found at here:

FACENET MANAGER
----------------

UNDER HEAVY DEVELOPMENT NOW

Roadmap: release facenet-manager on version 0.8

![asciicast](https://asciinema.org/a/113686?autoplay=1)
> The above ascii recording is just for demo purpose. Will replace it with facenet-manager later.

DOCUMENT

See auto generated docs

INSTALL & REQUIREMENT

`shell
$ npm install facenet
`

OS

Supported:
- [x] Linux
- [x] Mac
- [ ] Windows

Dependency

1. Node.js >= 7 (8 is recommend)
1. Tensorflow >= 1.2
1. Python3 >=3.5 (3.6 is recommend)

Make sure you run those commands under Ubuntu 17.04:

`shell
sudo apt install python3-pip
pip3 install setuptools --upgrade

`

Ram

| Neural Network Model | Task | Ram |
| --- | --- | --- |
| MTCNN | Facenet#align() | 100MB |
| Facenet | Facenet#embedding() | 2GB |

If you are dealing with very large images(like 3000x3000 pixels), there will need additional 1GB of memory.

So I believe that Facenet will need at least 2GB memory, and >=4GB is recommended.

API

Neural Network alone is not enough. It's Neural Network married with pre-trained model, married with easy to use APIs, that yield us the result that makes our APP sing.

Facenet is designed for bring the state-of-art neural network with bleeding-edge technology to full stack developers.

Facenet

`ts
import { Facenet } from 'facenet'

const facenet = new Facenet()
facenet.quit()
`

$3

Do face alignment for the image, return a list of faces.

$3

Get the embedding for a face.

`ts
face.embedding = await facenet.embedding(face)
`

Face

Get the 128 dim embedding vector for this face.(After alignment)

`ts
import { Face } from 'facenet'

console.log('bounding box:', face.boundingBox)
console.log('landmarks:', face.facialLandmark)
console.log('embedding:', face.embedding)
`

ENVIRONMENT VARIABLES

FACENET_MODEL

FaceNet neural network model files, set to other version of model as you like.

Default is set to models/ directory inside project directory. The pre-trained models is come from 20170512-110547, 0.992, MS-Celeb-1M, Inception ResNet v1, which will be download & save automatically by postinstall script.

`shell
$ pwd
/home/zixia/git/node-facenet

$ ls models/
20170512-110547.pb
model-20170512-110547.ckpt-250000.index
model-20170512-110547.ckpt-250000.data-00000-of-00001
model-20170512-110547.meta
`

DOCKER

![Docker Pulls](https://hub.docker.com/r/zixia/facenet/) ![Docker Stars](https://hub.docker.com/r/zixia/facenet/) ![Docker Layers](https://microbadger.com/#/images/zixia/facenet)

DEVELOP

![Issue Stats](http://issuestats.com/github/zixia/node-facenet) ![Issue Stats](http://issuestats.com/github/zixia/node-facenet) ![Coverage Status](https://coveralls.io/github/zixia/node-facenet?branch=master) ![Greenkeeper badge](https://greenkeeper.io/)

`shell
$ git clone git@github.com:zixia/node-facenet.git
$ cd facenet
$ npm install
$ npm test
`

COMMAND LINE INTERFACES




align

Draw a rectangle with five landmarks on all faces in the input\_image, save it to output\_image.

`shell
$ ./node_modules/.bin/ts-node bin/align.ts input_image output_image
`

embedding

Output the 128 dim embedding vector of the face image.

`shell
$ ./node_modules/.bin/ts-node bin/embedding.ts face_image
`




RESOURCES

Machine Learning

* Machine Learning is Fun! Part 4: Modern Face Recognition with Deep Learning
* Face recognition using Tensorflow
* Google: Our new system for recognizing faces is the best one ever
* A tensorflow implementation of "Deep Convolutional Generative Adversarial Networks
* What does Locality Sensitive Hashing Forests do? · maheshakya/my_first_project Wiki
* Average Face : OpenCV ( C++ / Python ) Tutorial

Python3

* Google Python Style Guide
* PyLint, PyChecker or PyFlakes?
* Useful Python Modules: Flake8
* PEP 8 - Style Guide for Python Code
* Python 3.6 venv — Creation of virtual environments

$3

* Mypy syntax cheat sheet (Python 3)
* Python 3 Type Hints and Static Analysis
* typing — Support for type hints

$3

* Stackoverflow: numpy-like package for node
* Read/manipulate/display images using NumJs
* Numjs - Like NumPy, in JavaScript
* ndarray - Modular multidimensional arrays for JavaScript

Dataset

1. LFW - Labeled Faces in the Wild

TODO

- [x] NPM Module: facenet
- [x] Docker Image: zixia/facenet
- [ ] Examples
- [x] API Usage Demo
- [x] Triple Distance Visulization Demo
- [ ] Performance Test(Align/Embedding/Batch)
- [ ] Validation Test(LFW Accuracy)
- [ ] Neural Network Models
- [x] Facenet
- [x] Mtcnn
- [ ] Batch Support
- [ ] ~~Python3 async & await~~
- [ ] Divide Different Neural Network to seprate class files(e.g. Facenet/Mtcnn)
- [x] K(?)NN Alghorithm Chinese Whispers
- [ ] TensorFlow Sereving
- [ ] OpenAPI Specification(Swagger)

INSPIRATION

This repository is heavily inspired by the following implementations:

* FaceNet by David Sandberg @davidsandberg
* OpenFace by CMU Satya Lab @cmusatyalab

CREDITS

1. Face alignment using MTCNN: Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks
1. Face embedding using FaceNet: FaceNet: A Unified Embedding for Face Recognition and Clustering
1. TensorFlow implementation of the face recognizer: Face recognition using Tensorflow

CONTRIBUTE

FaceNet Badge

![Powered by FaceNet](https://github.com/zixia/node-facenet)

`markdown
![Powered by FaceNet](https://github.com/zixia/node-facenet)
`

CHANGELOG

v0.9 master unstable

v0.8 (Apr 2018)

1. Added facenet-manager command line tool for demo/validate/sort photos
1. Switch to FlashStore npm module as key-value database

v0.3 Sep 2017

1. Added three cache classes: AlignmentCache & EmbeddingCache & FaceCache.
1. Added cache manager utilities: embedding-cache-manager & alignment-cache-manager & face-cache-manager
1. Added Dataset manager utility: lfw-manager (should be dataset-manager in future)
1. BREAKING CHANGE: Face class refactoring.

v0.2 Aug 2017 (BREAKING CHANGES)

1. Facenet#align() now accept a filename string as parameter.
1. BREAKING CHANGE: FaceImage class had been removed.
1. BREAKING CHANGE: Face class refactoring.

v0.1 Jul 2017

1. npm run demo to visuliaze the face alignment and distance(embedding) in a three people photo.
1. Facenet.align() to do face alignment
1. Facenet.embedding() to calculate the 128 dim feature vector of face
1. Initial workable version

TROUBLESHOOTING
---------------

$3

OS | Command
----- | -----
os x | brew install pkg-config cairo pango libpng jpeg giflib
ubuntu | sudo apt-get install libcairo2-dev libjpeg8-dev libpango1.0-dev libgif-dev build-essential g++
fedora | sudo yum install cairo cairo-devel cairomm-devel libjpeg-turbo-devel pango pango-devel pangomm pangomm-devel giflib-devel
solaris | pkgin install cairo pango pkg-config xproto renderproto kbproto xextproto
windows | instructions on our wiki

more os see node-canvas Wiki.

FAQ

1. facenet-manager display not right under Windows

See: Running Terminal Dashboards on Windows

2. Error when install: No package 'XXX' found

It's related with the NPM module canvas.

Error messages:
1. No package 'pixman-1' found
2. No package 'cairo' found
3. No package 'pangocairo' found

Solution for Ubuntu 17.04:
`shell
sudo apt install -y libpixman-1-dev
sudo apt-get install -y libcairo2-dev
sudo apt-get install -y libpango1.0-dev
`

Solution for Mac:
`shell
brew install python3
brew install pkg-config
brew install cairo
brew install pango
brew install libpng
brew install libjpeg
`

3. Error when install: fatal error: jpeglib.h: No such file or directory

It's related with the NPM module canvas.

Solution for Ubuntu 17.04:
`shell
sudo apt-get install -y libjpeg-dev
`

4. Error when run: Error: error while reading from input stream

It is related with the libjpeg package

Solution for Mac:
`
brew install libjpeg
`

5. Error when run:
`
Error: Cannot find module '../build/Release/canvas.node'
at Function.Module._resolveFilename (module.js:527:15)
at Function.Module._load (module.js:476:23)
at Module.require (module.js:568:17)
at require (internal/module.js:11:18)
at Object. (/Users/jiaruili/git/node-facenet/node_modules/canvas/lib/bindings.js:3:18)
at Module._compile (module.js:624:30)
at Object.Module._extensions..js (module.js:635:10)
at Module.load (module.js:545:32)
at tryModuleLoad (module.js:508:12)
at Function.Module._load (module.js:500:3)
`
It seems the package not installed in a right way, like sharp, canvas, remove the package and reinstall it.

run
`
rm -rf node node_modules/canvas
// if sharp, then remove sharp folder
npm install
`

6. Error when install
`
> facenet@0.3.19 postinstall:models /Users/jiaruili/git/rui/node-facenet
> set -e && if [ ! -d models ]; then mkdir models; fi && cd models && if [ ! -f model.tar.bz2 ]; then curl --location --output model.tar.bz2.tmp https://github.com/zixia/node-facenet/releases/download/v0.1.9/model-20170512.tar.bz2; mv model.tar.bz2.tmp model.tar.bz2; fi && tar jxvf model.tar.bz2 && cd -

x 20170512-110547.pb
x model-20170512-110547.ckpt-250000.data-00000-of-00001: (Empty error message)
tar: Error exit delayed from previous errors.
`

It seems this because not get the full model file successfully. See #issue63

Solution:

download the file from https://github.com/zixia/node-facenet/releases/download/v0.1.9/model-20170512.tar.bz2
rename the file model.tar.bz2 and move it to the folder models
try npm install` again

SEE ALSO
--------
1. Face Blinder: Assitant Bot for Whom is Suffering form Face Blindess
1. Wechaty Blinder: Face Blinder Bot Powered by Wechaty

AUTHOR

Huan LI \ (http://linkedin.com/in/zixia)

COPYRIGHT & LICENSE

* Code & Docs © 2017 Huan LI \
* Code released under the Apache-2.0 License
* Docs released under Creative Commons