Plumerault, M. Antoine (2018) Transfer learning for re-identification and classification on challenging datasets PFE - Project Graduation, ENSTA.
This document presents the work I have done during my internship at CEA fromthe 10th of April 2018 to the 9th October 2018. It describes several deep learn-ing techniques used to build a search engine for tattoos images based on visualcontent. We have explored several methods to produce a useful representation oftattoo images in which close points correspond to similar tattoos. The second partof the document is dedicated to my work on face recognition. To solve this task weused a large database of labeled images and explored different techniques to build arepresentation of face images that is useful for face recognition.
|Item Type:||Thesis (PFE - Project Graduation)|
|Uncontrolled Keywords:||Machine Learning, Deep Learning, Image Processing, Image Analysis,Face Recognition, Transfer learning, Metric Learning|
|Subjects:||Information and Communication Sciences and Technologies|
Mathematics and Applications
|Deposited By:||Antoine Plumerault|
|Deposited On:||10 avr. 2019 14:34|
|Dernière modification:||10 avr. 2019 14:34|
Repository Staff Only: item control page