Freelancer profile translated to English.
Back to original languageAbout Virgile
Engineer - specialized in machine learning with expertise in:
- Image and video processing with neural network methods and architectures for object detection, semantic and panoptic segmentation.
- Spatial information management with processing of satellite data, geolocated data (.tiff and geojson files), and remote sensing, also using neural networks, particularly with the torchgeo framework.
- Data analysis, scientific and matrix computation with numpy and scikit-learn.
- Global optimization of Python code. Study of spatial (memory usage), temporal (execution speed calculation) complexities and performance measurements.
**My technical skills**:
- Language: Python, Bash
- Deep Learning Framework: Pytorch, TorchGeo, Mmdetection, HuggingFace Transformers
- GIS and 3D/photogrammetry Software: QGIS, Agisoft Metashape
- Database: PostgreSQL (+PostGIS)
Looking forward to working with you!
French
Native or bilingual
English
Fluent
Remote only
Primarily works remotely
Experience
- Electric DreamDeep Learning EngineerVIDEO GAMES AND ANIMATIONSeptember 2022 - September 2022 (1 month)Paris, FranceTechnical support in a deep fake development project with *pytorch* targeting a person's face looking in a mirror:- Installation of CUDA, management of NVIDIA drivers and use of the CONDA environment manager.- Presentation of the latest github repositories introducing state-of-the-art deep learning models in this field.- Technical assistance in developing a solution for automatic face detection and bust cropping with mediapipe and opencv.- Generation of a *face-swap* type deepfake from a sample video.- Generation of a *face-reenactment* type deepfake from a sample video.- Demonstration of high-resolution deepfake image and video restoration (Real-ESRGAN, GFPGAN).
- ParticulierDeep Learning EngineerDEFENSE AND MILITARYDecember 2022 - December 2022Paris, France- Optimization of a panoptic segmentation model with pytorch and modification of a coco (JSON) image dataset.
- Image and Video Understanding Lab | IVUL - CEMSE | KAUSTResearch EngineerTECHApril 2021 - January 2022 (9 months)Thuwal, Saudi ArabiaArticle related to the mission: https://www.kaust.edu.sa/en/news/utilizing-artificial-intelligence-to-uncover-the-kingdoms-ancient-stone-structures- Development of an algorithm for large-scale automatic detection of archaeological structures in the Arabian Peninsula using the following object detection architectures: FasterRCNN, CenterNet (keypoints), and DETR coupled with satellite imagery.- Creation of a geolocated dataset with geojson files in *coco* format.- Development of an algorithm for a multi-scale approach to detections on satellite images with the addition of metadata and new visualization methods.- Creation and updating of databases with the generation of interactive maps.
Reviews
Recommendations
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- InternImage and Video Understanding Lab | IVUL - CEMSE | KAUST2022MSc Visiting Student à l'Image and Video Understanding Lab (IVUL) de la King Abdullah University of Science and Technology. - Développement des bases d'un algorithme pour la détection automatique à grande échelle de structures archéologiques dans la péninsule arabique en utilisant des méthodes d'apprentissage profond, des données satellitaires et la télédétection. - Développement d'un algorithme pour une approche multi-échelle avec l'ajout de métadonnées et de nouvelles méthodes de visualisation. méthodes de visualisation. - Création et mise à jour de bases de données avec les structures découvertes.
- Preparatory classes for Grandes Écoles PCSI/PSIInstitut Stanislas Cannes2015