Modeling real-world visual perception with deep learning

Modelling real-world visual perception with deep learning Deep neural networks (DNNs) have emerged as powerful computer algorithms that achieve human-level performance on challenging tasks such as object recognition and language comprehension. This has opened up a new research field in which explanations of brain function are sought in comparisons between DNNs and human behavior and brain activations. In this talk I will discuss why and how this computational modelling approach could indeed help us explain the brain, focusing specifically on the case of human visual perception of naturalistic real-world images and videos.