Saved in:
Bibliographic Details
Main Authors: Nguyen, Huy Hoang, Jung, Cédric, Salehi, Shirin, Glück, Tobias, Schmeink, Anke, Kugi, Andreas
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.23159
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917362798690304
author Nguyen, Huy Hoang
Jung, Cédric
Salehi, Shirin
Glück, Tobias
Schmeink, Anke
Kugi, Andreas
author_facet Nguyen, Huy Hoang
Jung, Cédric
Salehi, Shirin
Glück, Tobias
Schmeink, Anke
Kugi, Andreas
contents Foundation models for vision have transformed visual recognition with powerful pretrained representations and strong zero-shot capabilities, yet their potential for data-efficient learning remains largely untapped. Active Learning (AL) aims to minimize annotation costs by strategically selecting the most informative samples for labeling, but existing methods largely overlook the rich multimodal knowledge embedded in modern vision-language models (VLMs). We introduce Conformal Cross-Modal Acquisition (CCMA), a novel AL framework that bridges vision and language modalities through a teacher-student architecture. CCMA employs a pretrained VLM as a teacher to provide semantically grounded uncertainty estimates, conformally calibrated to guide sample selection for a vision-only student model. By integrating multimodal conformal scoring with diversity-aware selection strategies, CCMA achieves superior data efficiency across multiple benchmarks. Our approach consistently outperforms state-of-the-art AL baselines, demonstrating clear advantages over methods relying solely on uncertainty or diversity metrics.
format Preprint
id arxiv_https___arxiv_org_abs_2603_23159
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Conformal Cross-Modal Active Learning
Nguyen, Huy Hoang
Jung, Cédric
Salehi, Shirin
Glück, Tobias
Schmeink, Anke
Kugi, Andreas
Computer Vision and Pattern Recognition
Machine Learning
Foundation models for vision have transformed visual recognition with powerful pretrained representations and strong zero-shot capabilities, yet their potential for data-efficient learning remains largely untapped. Active Learning (AL) aims to minimize annotation costs by strategically selecting the most informative samples for labeling, but existing methods largely overlook the rich multimodal knowledge embedded in modern vision-language models (VLMs). We introduce Conformal Cross-Modal Acquisition (CCMA), a novel AL framework that bridges vision and language modalities through a teacher-student architecture. CCMA employs a pretrained VLM as a teacher to provide semantically grounded uncertainty estimates, conformally calibrated to guide sample selection for a vision-only student model. By integrating multimodal conformal scoring with diversity-aware selection strategies, CCMA achieves superior data efficiency across multiple benchmarks. Our approach consistently outperforms state-of-the-art AL baselines, demonstrating clear advantages over methods relying solely on uncertainty or diversity metrics.
title Conformal Cross-Modal Active Learning
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2603.23159