Bruno Cessac
Université Côte d’Azur, INRIA, France, INRIA Biovision team and Neuromod Institute - bruno.cessac@inria.frOur visual system is able to process our ever changing environment in a fast and efficient way. This processing starts at the retina, a very efficient organ, quite more complex than a simple camera. In particular, the retina is able to detect a moving object in a complex moving environment, to anticipate a trajectory, or even to signal violations in its expectations, with a very low energy consumption. I will show how these capacities can be explained by network effects and the propagation of waves induced by moving objects. Going through biophysical modeling and mathematical results I will discuss how these results could be insightful for the development of algorithms in AI and vision.
Bruno Cessac is a doctor in theoretical physics, research director at INRIA since 2010 and head of the Biovision team since January 2016.»His research was initially modeling and analysis of large sized dynamical systems arising in various fields such as physics, biology, sociology, computers networks. He has worked on subjects such as self-organized criticality, linear response in chaotic systems, social networks, communications networks. His main interest concerns neuronal networks dynamics. He has developed methods combining dynamical systems theory, statistical physics and ergodic theory allowing to classify dynamics arising in canonical neuronal networks models like integrate and fire models or firing rate model. He is now involved in developing models for the visual system, especially the retina, as well as numerical methods and software for neuroscientists. His current research is oriented toward 3 main goals: (i) Better understand how the modification of physiological parameters during development, addition of pharmacological substances, or pathologies induces changes in behavior in the retina and its functions; (ii) Model the effects of lateral connectivity – mainly amacrine cells – in retinal abilities such as motion anticipation; (iii) develop a realistic retino-thalamico-cortical model. All this theoretical work is done in collaboration with neuroscientists labs: Institut de la Vision, Institut des Neurosciences de la Timone, faculty of medical science of Newcastle, Center of Neuroscience of Valparaiso, Institut des Neurosciences Paris-Saclay .