Avatar

Alexandre Berthet

PhD Student - Image Processing and Deep Learning

Eurecom

Biography

Alexandre Berthet is currently PhD student in Digital Security at EURECOM – Sophia Antipolis since September 2019, with Professor Jean-Luc Dugelay as supervisor, working on DEFACTO project, whose subject is “Digital Image Forensics with Deep Learning”.

During his Master’s project, Alexandre worked in collaboration with EFREI Paris (Digital Engineers School) and Sorbonne University – Sciences Faculty, on the realization of an intelligent system for automated data collect on an apiary. The proof of concept made by Alexandre is integrated in the project projet « PNAPI » (Plateforme Numérique d’accompagnement des APIculteurs), laureate of CASDAR RT 2018.

During his first Master’s year, Alexandre worked in Laboratoire d’Informatique de Paris 6 (LIP6), in SoC department (Système On Chip) and more specifically in SYEL team (Systèmes Electronique). He contributed on SpinalCOM project, under Sylvain Feruglio’s direction via the realization of spectral discrimination’s methods of an optical sensor CMOS for the spinal cord’s monitoring.

Interests

  • Digital Image Forensics
  • Deep Learning
  • Camera Model Identification
  • Manipulation Detection
  • Forged Regions Localization

Education

  • PhD in Digital Image Forensics, now

    Eurecom

  • Master in Engineering Sciences, 2019

    Sorbonne University

  • Bachelor in Electronics, 2017

    Sorbonne University

Skills

Python

80%

Image Processing

80%

Deep Learning

90%

Experience

 
 
 
 
 

Data Scientist Junior

IBM France

Oct 2018 – Sep 2019 Paris

Missions :

  • Data Mining on mechanical parameters

  • Binary Text Classification with AI

  • IBM Deep Learning tools presentation

Skills : Python - keras / sklearn

 
 
 
 
 

Intern

Laboratoire d’Informatique Paris 6 (LIP6)

May 2018 – Jul 2018 Paris

Mission : Implementation of algorithmic method of spectral discrimination for the monitoring of the spinal cord

Skill : Matlab

Projects

*

DEFACTO Project

DEFACTO consortium (UTT, EURECOM and SURYS) participate to DEFALS Challenge. Automated detection of digital images falsifications.

PNAPI Project

Participation to the beginning of the PNAPI project. Proof of concept of an intelligent system for automated data collect on an apiary.

Recent Posts

Deep Reinforcement Learning : a review

Take few minutes to learn about basics in Reinforcement Learning and know more about methods of Deep Reinforcement Learning

Intelligent System for automated data collect on an apiary

This article explain the proof of concept realized for the PNAPI project.

Recent Publications

Two-stream Convolutional Neural Network for Image Source Social Network Identification

The identification of the source social network from an image is a relatively new research area in the image forensic domain. The …

A review of data preprocessing modules in digital image forensics methods using deep learning

Access to technologies like mobile phones contributes to the significant increase in the volume of digital visual data (images and …

Contact