Best Student Paper Award for Vanessa Borst et al. at the ADS Track of ECML PKDD 2025
09/18/2025The paper "WoundAmbit: Bridging State-of-the-Art Semantic Segmentation and Real-World Wound Care" by V. Borst, T. Dittus, T. Dege, A. Schmieder, and S. Kounev has received the Best Student Paper Award at the ADS Track of ECML PKDD 2025.
Vanessa Borst, Timo Dittus, Tassilo Dege, Astrid Schmieder, and Samuel Kounev won the Best Student Paper Award at the Applied Data Science (ADS) Track of ECML PKDD 2025.
Paper in a Nutshell
Chronic wounds affect millions of patients worldwide, particularly elderly and diabetic individuals. With WoundAmbit, the authors present an end-to-end solution for automated wound size estimation from RGB images, bridging the gap between modern deep learning and practical wound care. In a comprehensive benchmark, they evaluated state-of-the-art deep learning models for wound segmentation on both public datasets and real-world clinical images. Beyond segmentation quality, they assessed generalization, efficiency, and interpretability, and proposed a novel reference-object approach for clinically relevant wound size estimation. The results underscore the potential of modern vision architectures to advance telehealth solutions for remote wound monitoring.
Links
More information
- (German) press release: UKW
- More about ECML PKDD 2025: https://ecmlpkdd.org/2025/
Additional images
