Hilde Kuehne

Dr. Hilde Kuehne is a Professor at the University of Tuebingen and a key researcher at the Tuebingen AI Center. She holds affiliations with MIT-IBM Watson AI Lab and Goethe University Frankfurt, with a focus on computer vision, multimodal learning, and explainable AI. Her work bridges vision-language models, audio-visual alignment, and self-supervised methods.

• Co-organizer of the New Frontiers in Associative Memories workshop @ ICLR 2025
• Member of the Scientific Advisory Board of the Carl-Zeiss-Foundation
• Contributor to CVPR 2025's UTD dataset for unbiased video benchmarks

Her research addresses critical challenges in:

• Explainability for Vision Transformers (LeGrad)
• Fine-grained audio-visual alignment (CAV-MAE Sync)
• Zero-shot visual recognition automation (Meta-Prompting)
• Spatio-temporal grounding without annotations

Recent collaborative work spans 15+ publications across CVPR, NeurIPS, ICCV, and ICLR, with emphasis on multimodal foundation models, dataset bias mitigation, and differentiable logic networks. She actively contributes to workshop organization and peer review as evidenced by her involvement in CVPR 2025 and ICLR 2025 program committees.