David Tran-Thanh NGO

Sixense Mapping/Vinci Construction

R&D engineer

R&D Engineer in charge of computer vision/machine learning at ScanLab (an excellence center for Vinci Construction)

Specific tasks: development of a novel algorithm for automatic detection of accropodes from point cloud, development of deep learning-based defects detection (of civil structures such as building concrete crack, concrete spalling, ...) from orthophoto using Convolutional Neural Networks..

2019-now

CEA Cadarache/IRFM

R&D engineer

Working program: conception and development of a software platform for the exploitation of infrared thermography and visible diagnostics in nuclear fusion devices (WEST tokamak and W7-X stellarator).

Specific tasks: algorithm design (object detection, deep learning technique) and programming (in C++ with TensorFlow) for thermal event detection and tracking, development of the visual GUI application of the system (with Qt library).

2016-2019

Université de Strasbourg

R&D engineer

My PhD thesis is devoted to the detection of shadows, vegetation and buildings from single high resolution optical remote sensing images.

The first contribution of the thesis introduces a novel method that detects simultaneously shadows and vegetation from an high resolution optical image.

The second contribution of the thesis is a novel approach (ICIP paper and JSTARS paper) for the automated detection of buildings from monocular VHR optical images.

2012-2015

Télécom Physique Strasbourg

Teaching assistant

I have served as a teaching assistant during my PhD. I have experience in two modes of instruction: a) workshops for solving problems in classrooms and b) practice of programming using C and Matlab. Specifically, I have been in charge of four courses (for first- and second-year engineer students): image processing with Matlab programming, signal processing, numerical analysis and statistics.

2012-2015

Ifremer (Brest) - CLS (Toulouse)

Intern

My engineering intership was devoted to geolocate fish (european bar) from data storage tags. We presented a new hidden markov model that infers pelagic fish positions from the sole use of high-resolution temperature and depth histories. The project is detailed in this paper.

4/2012-9/2012