Skip to content

MeloniZippoProjects/MultimediaProject

Repository files navigation

Introduction

This is an Android app that performs face recognition and identification through the smartphone's camera. The app uses an internal database containing the known identities, including sample photos. The app allows the user to check the database content, delete or add new identities.

Implementation overview

  • Face detection using Haar Cascade Classifier
  • Face features exctracted using VGG2
  • Classification performed with weighted kNN

OpenCV library was used for data processing.

For more details, see the presentation.

Cannot find the caffe model?

If you get errors similar to Can't open "/data/user/0/melonizippo.org.facerecognition/cache/caffe_model/resnet50_ft_caffe/resnet50_ft.caffemodel", then you must first extract the 7z multipart file called resnet50_ft.7z.*** in the path src/main/assets/caffe_model/resnet50_ft_caffe/.

About

Android app to detect and recognize faces

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages