Neural Information Processing

My laboratory investigates human perception combining psychophysical experiments with computational modelling. Currently we have four research foci: First, to improve our image-based model of early spatial vision. Second, to connect early spatial vision with mid-level vision: perceived lightness, brightness and contrast in relation to surface reflectance and illumination in images of real scenes. Third, we investigate differences and similarities between deep convolutional neural networks and human object recognition. Fourth, we explore connections between causality from a perceptual as well as a machine learning perspective.