Overview of ImageCLEFMedical 2025 GANs Task: Training Data Analysis and Fingerprint Detection
| Art der Publikation: | Artikel |
| Zitat: | |
| Publication status: | Published |
| Zeitschrift: | CLEF 2025 Working Notes, 9 – 12 September 2025, Madrid, Spain / coeur-ws.org |
| Band: | 4038 |
| Jahr: | 2025 |
| Monat: | September |
| Seiten: | paper 169 |
| URL: | https://ceur-ws.org/Vol-4038/p... |
| Abriss: | The 2025 ImageCLEFmedical GANs Task - Controlling the Quality of Synthetic Medical Images created via GANs, continuing to investigate privacy and security concerns around using patient data to generate synthetic medical images. It comprises two complementary sub-tasks: the first extends prior editions by asking participants to detect which real images were used in training a Generative Adversarial Network to produce given synthetic outputs; the second builds on the 2024 findings by requiring teams to attribute each synthetic image to its specific real-image subset of origin. Ground-truth annotations and benchmark datasets of real and GAN-generated lung CT slices are provided for both tasks, and evaluation is based on Cohen’s Kappa for Subtask 1 and accuracy for Subtask 2. 14 teams submitted runs for Subtask1 and 4 teams submitted runs for Subtask 2, totaling 95 submitted runs that used a variety of methods. This paper presents an overview of the task setup, datasets, and evaluation metrics, and summarizes and discusses the approaches and results of the . |
| Schlagworte: | Deep Learning, Generative Adversarial Networks, generative models, ImageCLEF benchmarking lab, medical imaging, medical synthetic data |
| Autoren | |
| Hinzugefügt von: | [] |
| Gesamtbewertung: | 0 |
|
Anhänge
|
|
|
|
|
Notizen
|
|
|
|
|
|
Themen
|
|
|
|
|
