Brain decoding of visual imagery

Created a garment that embeds biosensors (PPG, GSR) and translates emotional states of the wearer as colorful light, sponsored by OpenBCIMotor Imagery classification

Design of 3 brain-computer interfaces : music generation from brainwaves, mask displaying emotional states and dream paintings. Visual object recognition from EEG signal (THINGS-EEG). Creation of a dataset of EEG recordings during sleep (REM stage). Supervised classification of dream content.Quasi real-time generation of 3D costumes in AR, strongly accepted to NeurIPS 2023 Creative AI Track.Presentation of my first author paper « Symbiotic Co-Creation with AI » at ICML 2023 workshop on AI & HCITechnical environment : CATIA, XGenerative Design, MNE-Python, PyTorch, Unicorn Hybrid Black, OpenVibe, VQGAN-CLIP

Emotion recognition and personality traits inference from EEG, ECG and EDA signal. Experiment design on affective ratings of EEG based Music Recommender Systems. Technical environment : Neurokit2, OpenBCI, BITalino

8-week course ”Digital Fabrication” with the Master 2 students of the program EIT Digital. Lectures and Practical Classes on Parametric Design.

Creation of perceptually-grounded metrics to evaluate the state-of-the art automatic lyrics-to-audio alignment models. First author of the paper "User-centered lyrics-to-audio alignment evaluation" accepted to Int. Soc. for Music Information Retrieval Conf. (ISMIR), 2021.Released my proposed evaluation metrics Perceptual-PCO in the mir_eval python library.

Development and deployment of a public karaoke application to run UX studies online. Design of two experiments to investigate the psychological validity of the current evaluation metrics used for the task of lyrics-to-audio alignment. Source separation and musical features extraction: beats, amplitude peaks, tempo and vowel duration.Technical environment : Docker, Kubernetes, FastAPI, Node.js, R

Mission for the Emotion and Perception Lab of Jagiellonian University. Development of the calibration and validation modules for the Pupil Labs eye-tracker in a VR headset (Vive Pro) and integration of these modules with the PsychoPy library to control the experimental procedures.Technical environment : Python, PsychoPy3, ZeroMQ, Pupil Core, OpenCV, HTC Vive Pro

Data Lake Analytics to enhance Ina STAT publications. Data modeling and datamart conception with Google BigQuery.Creation of transmedia indicators for the analysis of audiovisual programs.The dataset consisted of the metadata of all the french televisual archives since 1950.Design and full stack development of the REST API "Baromètre+" to automate and visualize reports in a dashboard. Complete a study on the media coverage of Charlie Hebdo, published in InaGlobal.Dimensional modeling, Business Intelligence, DataVizTechnical environment : PostgreSQL, BigQuery, D3.js, Gephi, Symfony

Mission for the Institute of Research and Innovation (IRI, Centre Pompidou).Analysis of "Plaine Commune, Contributing Learning Territory", a transformation project of a territory of 400,000 inhabitants, in collaboration with the philosopher Bernard Stiegler. Digital Transformation, Contributive Economy, SmartcitiesProject Manager of a team of 6 engineering studentsFunctional Analysis, Design Methodology

Achieving partnerships and visibility exchangeAccreditation Management1,000 spectators