Thibaut Chauffier

Senior ML Research Engineer
Computer Vision · Generative AI

Paris, France · Edition 2026

Senior ML Research Engineer bridging deep research and production impact — published author, inventor, and technical lead across research and engineering teams. 8+ years advancing generative models, 3D scene representations, and self-supervised vision systems, with rare end-to-end ownership: from novel architecture design to low-latency systems serving millions globally across robotics, UAVs, and beauty tech.

8+
Years
5
Patents
1M+
Users

Publications & Patents

2025

Locally Controlled Face Aging with Latent Diffusion Models

ICCV 2025 P13N Workshop · Oral Presentation · Co-author

Diffusion · Spatial Conditioning · Face Synthesis→ arXiv:2507.21600

5 Granted Patents

Spanning ML-driven robotics (active learning self-calibration), computer vision (skin tone detection & shade matching), and embedded AI inference systems.

Project Archive

001

Controllable Face Aging

Custom Latent Diffusion model with efficient spatial conditioning for real-time face aging on iPad Pro. Accepted as Oral at ICCV 2025.

Diffusion · PyTorch · CoreML · Edge Inference→ Publication
002

3D Facial Capture Pipeline

Novel 2D/3D Gaussian Splatting + FLAME pipeline for high-fidelity facial reconstruction, optimized for production deployment.

3D Gaussian Splatting · NeRF · FLAME · Computer Vision
003

Self-Supervised Facial Foundation Model

Foundation model trained with self-supervised objectives to learn semantically rich facial representations, enabling zero-shot transfer across multiple downstream tasks.

Self-Supervised Learning · Transfer Learning · PyTorch
004

Global Skin Tone Detection

Low-latency CNN-based system for skin tone detection and foundation shade matching, deployed across 9 L'Oréal brands on mobile and web. Sub-100ms inference via model serving pipeline.

Computer Vision · ONNX · FastAPI · MLOps→ eShadefinder
005

AI Formulation Engine

Real-time edge inference formulation prediction for YSL Rouge-Sur-Mesure, enabling customer-designed lipsticks under strict embedded constraints.

Embedded ML · Edge Inference · Real-time→ Rouge-Sur-Mesure

Employment Record

2022 — Present

Senior ML Research Engineer · Technical Lead

L'Oréal · Paris

  • Technical lead across AI and computer vision research and engineering teams
  • Designed and deployed low-latency ML systems serving millions of users globally
  • Led exploratory research (diffusion models, 3D Gaussian splatting, self-supervised learning) from prototyping to production
2017 — 2022

Mechatronics Engineer

L'Oréal · Paris

  • Designed ML-powered self-calibration system using Gaussian Process Regression with active learning — reduced material waste by 71% and downtime by 83%
  • Managed and mentored the embedded software team through full product lifecycle and safety certification
2016 — 2017

Mechatronics Engineer

Civic Drone · Paris

  • Developed real-time UAV fall detection using sensor fusion and Kalman filtering
  • Achieved 44% propulsion efficiency gain and 10% power reduction via ML-based optimization

Technical Capabilities

ML / AI Research

  • Diffusion Models
  • Gaussian Splatting · NeRF
  • Transformers
  • Self-Supervised Learning
  • Multimodal Models
  • LoRA / PEFT Fine-tuning
  • GANs

Frameworks

  • PyTorch
  • TensorFlow
  • HuggingFace (Transformers, Diffusers)
  • FastAPI
  • Docker

Production & MLOps

  • Model Serving
  • Edge Inference
  • ONNX
  • CUDA
  • GCP / Vertex AI
  • Quantization · Distillation
  • CI/CD

Education

2010 — 2015

Université de Technologie de Compiègne

Sorbonne University Alliance

Engineering Degree in System Engineering (MSc.)

Computer Science · Mathematics · Statistics · Physics · Mechatronics

Open Source