Assist. Prof. Dr. Sc. Simone Schaub-Meyer | TU Darmstadt

Abstract:
Deep learning has led to remarkable progress in computer vision, yet benchmark accuracy alone provides only a limited view of model capabilities. In this talk, Simone Schaub-Meyer argues that a deeper analysis of model behavior can yield both practical improvements and conceptual insights. First, she shows how fine-grained performance analyses can lead towards simpler and more computationally efficient vision models. In the second part, she focuses on understanding model behavior through visual explanations. She discusses the role of visual explanations in improving classification and presents a method to obtain such explanations efficiently in practice. She then highlights recent findings showing that deep networks rely not only on the presence of visual concepts, but also on their absence, and demonstrates how extending attribution and feature visualization methods makes these effects visible. Together, these perspectives illustrate how a deeper understanding of model behavior can inform the design of more efficient, reliable, and interpretable vision models.

Bio:
Simone Schaub-MeyerExternal link is an Assistant Professor at the Technical University of Darmstadt, where she leads a research group on image and video analysis. Her work focuses on developing efficient, robust, and interpretable methods for visual perception Her research is supported by the Emmy Noether Programme of the German Research Foundation. Assistant Professor Schaub-Meyer is a member of the ELLIS Unit Darmstadt and the Hessian Center for Artificial Intelligence (hessian.AI). She received her doctoral degree from ETH Zurich in collaboration with Disney Research Zurich, where her thesis on motion representation and video frame interpolation was awarded the ETH Medal.