Revisiting Active Learning in the Era of Vision Foundation Models

arXiv:2401.14555v2 Announce Type: replace-cross
Abstract: Foundation vision or vision-language models are trained on large unlabeled or noisy data and learn robust representations that can achieve impressive zero- or few-shot performance on diverse tasks. Given these properties, they are a natural fit for active learning (AL), which aims to maximize labeling efficiency. However, the full potential of foundation models has not been explored in the context of AL, specifically in the low-budget regime. In this work, we evaluate how foundation models …

active learning arxiv cs.cv cs.lg foundation replace type vision