Development of machine learning protoypes to classify, detect and localize threats in X-ray images.

- Feature extraction using high-level descriptors like HoGs and Wavelets.

- Feature classification using non linear classifiers (SVM, KNN) and ensemble methods (Gradient Boosting, Random Forest).

- Implementation of deep learning methods to classify images (CNN) and localize threats or objects for recognition purpose (RCNN).

- Adding Semantic information (WordNet) to develop Content-Based Image Retrieval system for cargo inspection.

- Assisting software team for release implementations.

Development of machine learning functions on color images to analyse (colour, shape, license plate) and control vehicles at frontiers.

- Localize vehicle using high level image processing (Hough transform).

- Develop SIFT feature extractor for shape recognition.

- Colour vehicle classfication using a derivation of PLSA dataminig algorithm (Probabilistic Latent Semantic Analysis) and color hierarchical clustering (Mean-Shift algorithm).

Development of a prototype function on ground pulse radar, to detect exact positions and speeds targets (solving well-known distance/speed ambiguities)

- Clustering elementary detections using Hierarchical Ascendant Clustering with Ward distance.

- Prune false alarm by satistics test (goodness-of-fit) on cluster compactness.

- Development led on synthesized data and fine-tuning parameters led on ground-truth recordings.

- Assisting sofware team for release implementation and validation team for real time analysis.

Prototype an antena mesh and simulate it to meet performance requirements (central frequency and bandwith, spatial directivity).

- Vapnik-Chervonenkis theory for statistic learning

- Supervised/Unsupervised Learning

- Markov Chains and Bayesian Networks

- NoSQL database

- Storage and Computation (Map/Reduce, PageRank) of distributed data

- Data vizualisation

- Deterministics signals and frequency analysis.

- Random signals and statistics.

- Image processing for computer vision.

Scientific preparation to the french Grandes Ecoles with competitive entrance examinations.

  Python

  Matlab

  Lua

  C/C++

  Java

  Open Computer Vision

  libSVM

  Scikit-Learn

  XgBoost

  MatConvnet

  Keras

  Torch

  Hadoop (HDFS,MapReduce)

  Apache Spark

  bash

  Git

  Jekyll

  Markdown

  Latex

Work as waiter in London city centre

Provide mathematics & physics courses to a college student

Movies (Bollywood, Japanimation), travelling, daily fitness, cooking, French literature.