Aim and scope

In recent years, we have witnessed unprecedented advancements in the automatic analysis of visual data by computer algorithms, thanks to a series of factors which unleashed the potential of convolutional neural networks. Part of the reason of their success could be explained by the biologically-inspired design of such models: indeed, while the classic artificial neuron may only by an extreme simplification of the biological neuron, the increasing representational complexity learned by CNNs may be more faithful to the layered structure of the lower areas of the human visual cortex. However, the road towards achieving a degree of artificial emulation of the visual system high enough to interpret an environment as humans do is still long: while we are able to identify low- to high-level visual patterns from images and videos, artificial models largely miss the human capability to make sense of this information, recognize semantic patterns, correlate to memory and experience, and so on. The proposed workshop is inspired by the realization that understanding and speculating on the yet mostly unknown mechanisms used by the brain to process visual information and knowledge may be the key to further advance computer vision beyond the black-box paradigm of training from data in the hope to uncover the very same processes. The daunting yet fascinating challenge presented by this task calls for a largely multidisciplinary effort by research communities in the fields of artificial intelligence, machine learning, cognitive neuroscience, psychology, among others. The aim of the workshop is to excite the study and development of paradigms, methods and tools for computer vision driven or inspired by the human brain, both as a computational model and a source of data, and to promote the diffusion of new benchmarks and evaluation protocols to support the scientific community in the pursuing of a better understanding of the brain processes underlying human visual perception and comprehension.

Topics




Workshop Program

Day: September 8th, afternoon
Schedule:

13:30 Workshop starts

13:45- 14:30 Keynote (John Tsotsos, York University)

14:30- 16:00 Poster session

16:00- 16:30 Coffee break

16:30- 17:15 Keynote (Concetto Spampinato, University of Catania)

17:15- 17:45 Discussion

17:45- 18:00 Conclusions and future directions

18:00 Workshop ends


Organizers