Model the living structures:
Neuroscience, joint progress with algorithmics

January 2005

How image processing, scientific computing and brain function modeling methods help us advance in understanding the way information is processed and aggregated, solve complex problems and transform a simple observation into durable learning? Various INRIA teams contribute to this common effort thanks to close collaborations with neurophysiologists.

Visualization of the main connection paths between different areas of the cortex

Biological computing and algorithmic computing-there seems to be a natural partnership between neuroscience studying the brain and nervous system of living beings on the one hand, and computing and mathematics on the other hand. The development of computer science was linked to that of machines that compute faster than the human brain, when it comes to carrying out simple tasks. There is no artificial intelligence today capable of attaining the flexibility and robustness of the human brain and nervous system when it comes to seeing, moving or recognizing an object. The study of brain functions is a powerful source of inspiration for algorithmics that opens up new directions in computing-distributed computing, emergent computing, and adaptive computing are a few examples. Simultaneously, as in other life sciences, modeling, scientific computing and computer science are indispensable partners to organize and process the data acquired during experiments, analyze behaviors and model functions.

This is why more and more research projects are created that associate neuroscience, computing and mathematics at different levels of analysis, from the infinitely small-the molecule and the neuron observed via the signals supplied by cerebral imaging-to global behavior studied by integrative neuroscience.

CORTEX models neuron computing on several levels, from the neuron to the reproduction of certain simple functions. The models are then implemented in robots.

The goal of the research carried out by ODYSSEE is to better understand vision by associating the study of biological vision with the development of algorithmic vision tools.

OMEGA is a team specializing in probability theory. It applies its research to the study of the electrical activity of neurons and its importance in information exchange.Advances in this research also contribute to progress in very varied disciplines such as robotics, medicine, ergonomy, learning, telecommunications, financial mathematics, and so on.

There are considerable developments in artificial intelligence concerning computing and perception (vision, olfaction, etc.). Some scientists are starting to explore domains that were long considered to be inaccessible such as emotion, via research on selective attention, for example, motivation, learning or even consciousness.

• CORTEX: Neuromimetic intelligence
  Can we model and reproduce neuron computing, that is to say the functioning of the brain?
• ODYSSÉE: Algorithmic and biological vision
  Will progress in medical imaging make it possible to understand and model human or animal vision mechanisms in all their complexity?
• OMEGA: Probabilistic numerical methods
  Will probability theory make it possible to better understand the electrical activity of neurons?

Contact :

INRIA Vincent Coronini

Tel : +33 01 39 63 57 29