RubyFaceRecognition

A Ruby wrapper for recognize and manipulate faces.

Enables

• Fetch face encoding.
• Get distance between two face encodings.
• Compare two faces are same person.

Dependencies

• pycall >= 1.4.1, see https://github.com/mrkn/pycall.rb for more.

• python 3

  • Install python 3.7 and face recognition

    1. If use pyenv, install python with command to load libpython in pycall:

       envPYTHON_CONFIGURE_OPTS="--enable-framework" CFLAGS="-I$(xcrun --show-sdk-path)/usr/include" pyenv install 3.7.0
       

    2. Export LIBPYTHON environment variable:

       export LIBPYTHON="/Users/xxx/.pyenv/versions/3.7.0/lib/libpython3.7m.a"
       

    3. Install face recognition tool, see https://github.com/ageitgey/face_recognition:

       pip3 install face_recognition
       

  • Another way for ubuntu

    1. Install python3 and pip3

    2. Install dlib

       git clone https://github.com/davisking/dlib.git
       

       cd dlib
       

       mkdir build; cd build; cmake ..; cmake --build .
       

       cd ..
       

       python3 setup.py install
       

    3. Install plugin

       pip3 install face_recognition
       

• face_recognition, see https://github.com/ageitgey/face_recognition for more.

Installation

Add this line to your application's Gemfile:

gem 'ruby_face_recognition'

And then execute:

$ bundle install

Or install it yourself as:

$ gem install ruby_face_recognition

Usage

  require 'ruby_face_recognition'

  # See ./examples folder for images
  wdz1_doubles = FaceRecognition.fetch_face_encoding('./wdz1.jpeg')
  wdz2_doubles = FaceRecognition.fetch_face_encoding('./wdz2.jpeg')
  ycg1_doubles = FaceRecognition.fetch_face_encoding('./ycg1.jpeg')
  ycg2_doubles = FaceRecognition.fetch_face_encoding('./ycg2.jpeg')

  pp "wdz1 and wdz2 is same: #{FaceRecognition.same_person?(wdz1_doubles,
                                                          wdz2_doubles)}, tolerant is #{FaceRecognition.get_tolerant(
                                                              wdz1_doubles, wdz2_doubles
                                                          )}"
  pp "wdz1 and ycg1 is same: #{FaceRecognition.same_person?(wdz1_doubles,
                                                          ycg1_doubles)}, tolerant is #{FaceRecognition.get_tolerant(
                                                              wdz1_doubles, ycg1_doubles
                                                          )}"
  pp "ycg1 and ycg2 is same: #{FaceRecognition.same_person?(ycg1_doubles,
                                                          ycg2_doubles)}, tolerant is #{FaceRecognition.get_tolerant(
                                                              ycg1_doubles, ycg2_doubles
                                                          )}"

  #Output:
  "wdz1 and wdz2 is same: true, tolerant is 0.31643992115339953"
  "wdz1 and ycg1 is same: false, tolerant is 1.1141872408732252"
  "ycg1 and ycg2 is same: true, tolerant is 0.338220706580959"

Use in Rails

Assuming Database is postgresql:

• Create column to save face encoding.

  add_column :table_name, :column_name, :float, array: true, comment: 'face encoding'
  

• Create calculating distance function script to Database.

    class AddDistanceFunctionToDb < ActiveRecord::Migration[6.1]
        def self.up
            execute "CREATE OR REPLACE FUNCTION distance(l double precision[], r double precision[]) RETURNS double precision AS $$
            DECLARE
                s double precision;
            BEGIN
                s := 0;
                FOR i IN 1..128 LOOP
                s := s + power(l[i] - r[i], 2);
                END LOOP;
                RETURN sqrt(s);
            END;
            $$ LANGUAGE plpgsql;"
        end
        def self.down
            execute "drop function distance(l double precision[], r double precision[]) cascade"
        end
    end

• Use distance function in SQL

  tolerant = ActiveRecord::Base.connection.exec_query("select distance(ARRAY#{face_encoding1}, ARRAY#{face_encoding2})").rows[0][0]

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install. To release a new version, update the version number in version.rb, and then run bundle exec rake release, which will create a git tag for the version, push git commits and tags, and push the .gem file to rubygems.org.

Contributing

Bug reports and pull requests are welcome on GitHub at https://github.com/[USERNAME]/ruby_face_recognition. This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the code of conduct.

License

The gem is available as open source under the terms of the MIT License.

Code of Conduct

Everyone interacting in the RubyFaceRecognition project's codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.