Face Similarity Search

Problem Statement

Status: available
Supervisor: Martin Kampel

Start: as soon as possible

Facial analysis, face recognition or facial similarity search has been a very well-known topic in computer vision for many, many years. Today we have mobile solutions as well as large datasets for comparing and analysing face images with varying accuracy. Input data might be live streams, photos, surveillance videos, historic footages or others.

Goal

The goal of this project is to study current state of the art and existing solutions for face similarity search. Within the project one running solution should be implemented and finally provided. Next to functionality two important aspect could be mentioned: that is practicability and evaluation.

If you are interested, please contact Martin Kampel

References

https://github.com/sthalles/face-similarity
https://github.com/harveyslash/Facial-Similarity-with-Siamese-Networks-in-Pytorch

https://hackernoon.com/facial-similarity-with-siamese-networks-in-pytorch-9642aa9db2f7
https://www.freecodecamp.org/news/how-to-train-your-own-faceid-cnn-using-tensorflow-eager-execution-6905afe4fd5a/

Zagoruyko, Sergey & Komodakis, Nikos. (2015). Learning to compare image patches via convolutional neural networks. 4353-4361. 10.1109/CVPR.2015.7299064.

Koch, Gregory R.. “Siamese Neural Networks for One-Shot Image Recognition.” (2015).