|
Evènement pour le groupe Séminaire des Doctorants
Date | 2013-11-28 13:30-14:30 |
Titre | [Semidoc] Modeling the early visual system: The story of the Olivier salad |
Résumé | The secret recipe of how the visual system works has been one of the utmost sought discoveries in neuroscience during the last decades. Considering that it uses around 50% of the resources of the brain (directly or indirectly), the massive initiatives studying the brain nowadays are expected to help unraveling also this mystery. However, there is much more than the brain in the visual system. For instance, the retina not only transduces light, and the thalamus not only relays information: the early visual system is a complex network of sub-networks interacting with each other. This complexity has resulted in expensive computational models with many parameters. Here, parallelizing with the story of the Olivier salad, we show how a functional recipe can include both new knowledge and hypotheses to explain visual treatment, phenomena, and their links to cognition and behavior, using simple bio-inspired algorithmic implementations that do not require the complexity of existent computational models.
This presentation starts with the classical recipe of human vision, to explain what this view is missing, and how we try to understand such distributed dynamical asynchronous system. Then, it develops the difficulties in the process and how, if we focus on functionality, we can implement a cheaper and simpler systemic computational model that allows us to bind knowledge and hypotheses, so cognition and behavior can be studied without losing taste. |
Lieu | Amphi du LaBRI |
Orateur | Carlos Carvajal |
Url | INRIA, Equipe Mnemosyne |
Aucun document lié à cet événement. RetourRetour à l'index
| |