Seminars

My seminar presentations and paper discussions at Michelin and other venues.

Research Seminar Founder & Organizer

A Paper A Week (APAW)

Bi-Monthly Seminar

I founded and organize a bi-monthly research seminar that unites researchers and data scientists at Michelin.
The seminar promotes knowledge sharing and collaboration, highlighting state-of-the-art papers across several domains and topics, including: machine learning (ML), computer vision (CV), natural language processing (NLP), and uncertainty quantification (UQ) with practical applications to tire manufacturing.

The seminar provides a dynamic platform for researchers to present cutting-edge work, discuss recent advances, and explore pathways for translating academic research into industrial applications. By bringing together diverse perspectives across the organization, it fosters interdisciplinary collaboration and accelerates innovation in data-driven manufacturing processes.
Serving as a key forum for knowledge exchange, the seminar helps researchers stay abreast of the latest developments while identifying opportunities for practical implementation. Each session focuses on a single paper or topic, delivering presentations that seamlessly bridge theoretical insights and real-world applications, strengthening collaboration across teams and driving impactful solutions.

Date: April 2024 - Present
Location: Clermont-Ferrand
Audience: Researchers & Data Scientists

Frequency: Bi-monthly
Focus: ML, CV, NLP, UQ
Format: Paper presentations & discussions

 

My Selected Paper Presentations

Quality-Diversity Optimization

arXiv (2020) · Evolutionary Algorithms · Optimization · Machine Learning

Explored quality-diversity algorithms for discovering diverse high-performing solutions, with applications to optimization problems in manufacturing.

Paper
Attention Is All You Need

NeurIPS (2017) · Deep Learning · Transformers

Discussed the transformer architecture and its revolutionary impact on sequence modeling, with potential applications to manufacturing processes.

Paper
HyenaDNA: Long-Range Genomic Sequence Modeling

arXiv (2023) · Sequence Modeling · Genomics · BioAI

Presented advances in long-range sequence modeling using Hyena operators, exploring connections to industrial time-series prediction tasks.

Paper
A Gentle Introduction to Conformal Prediction

arXiv (2021) · Uncertainty Quantification · Machine Learning

Introduced conformal prediction methods for providing reliable uncertainty estimates, crucial for quality monitoring in manufacturing processes.

Paper
TorchSOM: PyTorch Library for Self-Organizing Maps

GitHub Project · Unsupervised Learning · Machine Learning · Dimensionality Reduction

Presented my open-source PyTorch implementation of Self-Organizing Maps, demonstrating applications to industrial data visualization and anomaly detection.

GitHub Personal Package

 


 

Invited Seminars

Online Sensing for Quality Monitoring

July 2025
Data Scientists Network, Michelin
Quality Control · Real-time Monitoring

Proposed multiple online strategies for quality monitoring in the tire manufacturing process, focusing on real-time detection and adaptive methods for maintaining production quality.

TorchSOM: Applications to Online Sensing and Uncertainty Quantification

October 2025
Mathematical PhD Seminar, École polytechnique
Self-Organizing Maps · Online Sensing · Uncertainty Quantification

Presented TorchSOM, my open-source PyTorch implementation of Self-Organizing Maps, along with applications to online sensing and online uncertainty quantification in industrial contexts.

GitHub Personal Package