The goal of the Human-inspired Computer Vision workshop is to link and disseminate parallel findings in the fields of neuroscience, psychology, cognitive science and computer vision, to inform the development of human-inspired computational models, capable of solving visual tasks in a human-like fashion. 

Although pure engineering approaches to computer vision can achieve high performance in standard vision datasets regardless of their resemblance to human biological mechanisms, they still fail in many visual cognitive tasks that are fundamental to human intelligence, such as zero-shot visual search, context aware object recognition, and continual learning. We believe that expanding the perspective from a merely performance-oriented approach to the development of models holding human-like properties, while still achieving high performance, can push forward the frontiers of computer vision research. Indeed, on one hand, insights from cognitive science and neuroscience can help us identifying and tackling gaps between humans and machines. Characterizing aspects where humans still surpass computer vision, such us robustness to biases, working memory tasks, and visual search, can ultimately improve models’ performance as well. On the other hand, inspiration from biology can lead to computer vision models that are useful to explain neuroscientific and cognitive observations, and to deepen our understanding of human brain and developmental mechanisms. Such a perspective can pave the way to applications of computer vision outside classic domains, for instance to augment human vision and cognition in presence of neurodevelopmental disorders.