Cost-efficient Active Learning for Referring Image Segmentation and Grounding
2026-08-31 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial Intelligence
AI summaryⓘ
The authors address a problem in teaching computers to identify objects in images from natural language descriptions. They created a system that helps pick the best images to label, even when no text descriptions are available initially. By using AI models to guess labels and measuring confusion about which object a description refers to, their method selects images that are more useful for learning. They also designed an easy tool for people to write clear, focused descriptions faster. Their tests show this approach works better and speeds up labeling compared to other methods.
visual groundingactive learningreferring expressionsregion annotationsfoundation modelsacquisition functionambiguityimage labelingmachine learningannotation interface
Authors
Junbeom Hong, Seonghoon Yu, Hyung Rok Jung, Sundong Kim, Jeany Son
Abstract
Collecting natural-language referring expressions along with region annotations, such as masks or boxes, is a major bottleneck in visual grounding (VG), as annotators must write descriptions that distinguish target regions from visually similar ones. We tackle this by formulating active learning (AL) for VG under the realistic setting where only raw images are available without accompanying text. Since ground-truth text is unavailable, sample selection must estimate which images contain ambiguous regions that would require discriminative referring expressions. To address this, we generate auxiliary region-text pairs using foundation models, and introduce Referred Region Ambiguity, a new acquisition function that measures whether the model's confidence collapses onto a single region or disperses across multiple candidates. It allows our method to prioritize images with strong cross-region competition, which are more informative due to their visual ambiguity. We also design a referring-expression annotation interface that helps annotators quickly focus on writing discriminative language with a few clicks. Experiments on RIS and REC benchmarks show that our AL framework consistently outperforms several AL baselines, while a user study shows up to 1.6X faster description labeling of ours.