Quickly Tuning Foundation Models for Image Segmentation

Breenda Das, Lennart Purucker, Timur Carstensen, Frank Hutter

Submission Type: Short paper

Tldr: A meta-learning-based method (QTT-SEG) that rapidly fine-tunes SAM for domain-specific image segmentation with minimal effort, outperforming strong AutoML baselines.

Submission Number: 8