Laurent Rodriguez

Title: Laurent Rodriguez: Brain-Inspired Multimodal Classification

Introduction

Laurent Rodriguez is an inventor associated with Grasse, France. His work includes a patent for a computer-implemented approach to multimodal data classification using brain-inspired unsupervised learning. The patent also describes neuromorphic computing hardware designed to implement the method.

Latest Patents

Rodriguez is associated with 1 patent: “Method and System for Multimodal Classification Based on Brain-Inspired Unsupervised Learning.” The method trains multiple artificial neural networks on multimodal data, then learns associations between them through bidirectional lateral connections linking co-activated Best Matching Units. It labels neurons using a divergence algorithm and selects a global Best Matching Unit using a convergence algorithm.

Career Highlights

Rodriguez has been associated with the Centre National de la Recherche Scientifique and Université Côte d’Azur. His patent describes a framework for connecting neural networks trained on different modalities, with a neuromorphic hardware structure intended to support its implementation.

Collaborations

His listed coworkers include Lyes Khacef and Benoît Miramond.

Conclusion

Laurent Rodriguez’s patent focuses on combining unsupervised learning, multimodal neural-network associations, and neuromorphic hardware for data classification.