What are the computations underlying primate versus machine vision?
Primates excel at object recognition: For decades, the speed and accuracy of their visual system have remained unmatched by computer algorithms. But recent advances in Deep Convolutional Networks (DCNs) have led to vision systems that are starting to rival human decisions. A growing body of work also suggests that this recent surge in accuracy is accompanied by a concomitant improvement in our ability to account for neural data in higher areas of the primate visual cortex. Overall, DCNs have become de facto computational models of visual recognition.
In this talk, I will review recent work by our group which brings into relief limitations of modern DCNs as computational models of primate vision. I will show that visual features learned by DCNs from large-scale object recognition databases differ markedly from those used by human observers during visual recognition. I will further demonstrate that DCNs are limited in their ability to solve seemingly simple visual reasoning problems involving incremental grouping, similarity and spatial relation judgments suggesting the need for additional neural computations beyond those implemented in current architectures. I will show how neuroscience principles may help guide the future design for more robust computer vision architectures.
Dr. Serre is Associate Professor in Cognitive Linguistic & Psychological Sciences at Brown University. He received a Ph.D. in Neuroscience from MIT in 2006 and an MSc in EECS from Télécom Bretagne (France) in 2000. Dr Serre is Faculty Director of the Center for Computation and Visualization and the Associate Director of Carney’s behavioral core and the “SmartPlayroom”. Dr Serre has served as an area chair for machine learning and computer vision conferences including CVPR and NIPS. He is currently serving as a domain expert for IARPA’s Machine Intelligence from Cortical Networks (MICrONS) program and as scientific advisor for Vium, Inc. He is the recipient of an NSF early Career award as well as DARPA’s Young Faculty Award and Director’s Award. His research seeks to understand the neural computations supporting visual perception and has been featured in the BBC series “Visions from the Future” and appeared in several news articles (The Economist, New Scientist, Scientific American, IEEE Computing in Science and Technology, Technology Review and Slashdot).