Heiko SchüttView profile
Associate Professor
Heiko Schütt is Associate Professor for Computational Cognitive Science and Modeling at the Université du Luxembourg. His research focuses on developing mechanistic models of visual perception and cognition using deep neural networks, Bayesian inference, and efficient coding principles. He is also the developer of widely used toolboxes such as rsatoolbox for representational similarity analysis and Psignifit 4 for psychometric function fitting. His work lies at the intersection of cognitive science, computational neuroscience, and machine learning, with a strong emphasis on creating and evaluating models of human perception and decision-making. He investigates early visual processing, eye movement dynamics, and model evaluation methodologies, contributing both theoretical frameworks and practical tools to the scientific community. The recent publications reflect a strong trend toward developing rigorous statistical methods for comparing neural and cognitive models, particularly in the context of representational geometries and perceptual decision-making. His work increasingly bridges human cognition and artificial neural networks, exploring parallels in generalization and representation. Much of his research is image-computable and grounded in empirical psychophysics. He has previously held postdoctoral positions with Weiji Ma at New York University and Niko Kriegeskorte at the Zuckerman Institute, Columbia University. His PhD was jointly conducted with Felix Wichmann at Tübingen and Ralf Engbert at Potsdam, focusing on early visual processing and eye movements. No scientific awards are mentioned in the provided text. While no formal advisees are listed, his active research program and development of major scientific toolboxes suggest a role in mentoring students and collaborators. There is no mention of specific grants, but his work likely involves funded research given the scale and impact of his projects. He maintains active code repositories on GitHub for rsatoolbox, Psignifit 4, early vision models, and eye movement modeling, indicating leadership in open science and computational tool development. These resources support a broad community in cognitive and computational neuroscience.
