Work Experience

Software Implementation Specialist, Esker

— Lyon, France (Nov 2014 — Present)

  • Work as part of a team of 4 persons, or independently, to implement Esker’s on-premise software and SaaS solutions
  • Collaborate closely with B2B customers to define requirements, customize and test the application, and train end users (audience of 5 to 15 people)
  • Improve customers experience and resolve issues for clients by working with cross-organizational teams (sales, research & development and technical support)

Software Engineer, Worldline, an Atos company

— Villeurbanne, France (June 2013 — Dec 2013)

  • Designed and programmed, as part of a team, a web application using Java EE of an e-commerce website for the world's largest chain of fast food restaurants
  • Worked with a cross-functional team of front-end and back-end software engineers, mobile software engineers, and database administrators

Education

Nanodegree Program, Self-Driving Cars Engineer

— Udacity (2017)

Nanodegree Program, Machine Learning Engineer

— Udacity (2017)

Master's Degree, Telecommunications, Services and Computer Systems Networking

— Institut National des Sciences Appliquées de Lyon (2009 - 2014)

Master's Degree, Telecommunications, Services and Computer Systems Networking

— Exchange student, Shanghai Jiao Tong University (2014)

Master's Degree, Computer Science

— Exchange student, Politecnico di Torino (2012 - 2013)

Projects

Vehicle Detection and Tracking

See Project

  • Performed features extraction using YCrCb color transform, histograms of color, and Histogram of Oriented Gradients (HOG) on a labeled training set of images
  • Trained a Linear SVM classifier and proudly achieved a test accuracy of 98.8%
  • Implemented a sliding-window technique and used the trained classifier to search for vehicles in images
  • Successfully estimated a bounding box (using heat map) for tracking detected vehicles along the test video

Advanced Lane Finding

See Project

  • Implemented a software pipeline to detect lane lines in a video from a front-facing camera on a car, using advanced computer vision techniques (camera calibration, perspective transformation, sliding windows)
  • Successfully found lane lines from the test video

Behavioral Cloning

See Project

  • Collected data of good driving behavior from 3 cameras (front, left, right) using Udacity simulator
  • Designed, trained and validated a model based on the Nvidia DNN that predicts a steering angle from image data
  • Successfully drove the vehicle autonomously around a track in the simulator using the model

Traffic Sign Classification for Self-Driving Car

See Project

  • Implemented a deep neural network classifier (7 layers) which was inspired by LeNet-5: 4 convolutional layers for feature extraction and 3 fully connected layer for classification
  • Successfully boosted the training accuracy to 0.999 on the German traffic signs dataset using preprocessing techniques (histogram equalization, grayscaling, and normalization)

Lane Finding on the Road

See Project

  • Implemented a software pipeline to detect lane lines in a video from a front-facing camera on a car, using Python and OpenCV
  • Successfully found lane lines from 3 different test videos