Full Professor, University of Technology Nuremberg
2 papers at NeurIPS 2025
We propose a method to enable elastic inference for pretrained ViTs via structured pruning. We do not require labels, generalize to models without a classification head and do not need re-training. We improve over the existing state of the art.
DISCOVR is a self-supervised framework for echocardiography video representation learning that integrates spatial and temporal modeling, achieving strong generalization in anomaly detection, segmentation, and LVEF prediction.
