Computer Vision News - May 2023

34 Best Paper ISBI 2023 biomedical image domain, where you have small datasets that require expert annotation, an automated measurement like ours can open the doors for people with few resources to use other datasets and build up their projects. ” Away from this paper, Marcel is working at KIT on an interdisciplinary project about personalized oncology . Miguel has worked on tasks such as license plate and gas leakage detection, but his research focuses on medical and biomedical problems. In that regard, his PhD thesis is devoted to behavioral cell analysis . Taking home the Best Paper Award at ISBI 2023 last month is still sinking in for both, so is it too soon to ask what factors they think influenced the judges’ decision to medical problems, and it has been tested on multiple datasets, including histology , dermatology , and real-world images . Marcel and Miguel intend to build a latent space that includes all the standardmedical imaging modalities, including ultrasound, CT, and MRI. They will add more datasets and improve the architecture to solve the problem more robustly. “ We’ll continue working on this topic, and when this task of selecting the best available database is solved, it will create huge potential in the application of deep learning, ”Marcel says. “ Onceyouovercome the issue of finding annotated datasets, there will be plenty of future work in this direction. ” Miguel agrees: “ Yes, particularly in the B E PAP I S 2 0 IEEE ISBI