Hard to See, Hard to Label: Generative and Symbolic Acquisition for Subtle Visual Phenomena

TL;DR

This paper introduces GSAL, an active learning framework that combines diffusion-based difficulty signals with semantic coverage to improve detection of subtle, rare visual anomalies in industrial and natural datasets.

Contribution

GSAL uniquely integrates diffusion-based visual difficulty scoring with a hierarchical semantic coverage prior for more effective active learning in object detection.

Findings

GSAL outperforms uncertainty-based methods in rare class detection.

GSAL achieves higher label efficiency in defect inspection tasks.

Experiments show consistent improvements across multiple datasets.

Abstract

Subtle visual anomalies such as hairline cracks, sub-millimeter voids, and low-contrast inclusions are structurally atypical yet visually ambiguous, making them both difficult to annotate and easy to overlook during active learning. Standard acquisition heuristics based on discriminative uncertainty or feature diversity often overselect dominant patterns while underexploring sparse yet important regions of the data space. This failure mode is especially severe in industrial defect inspection, where anomalies may be both low-prevalence and difficult to distinguish from surrounding structure. To resolve this, we propose GSAL, an active learning framework for object detection that combines a diffusion-based difficulty signal with a hierarchical semantic coverage prior. The diffusion component scores images and proposals using reconstruction discrepancy and denoising variability, prioritizing visually atypical or ambiguous examples. However, diffusion alone does not prevent acquisition from repeatedly favoring hard samples within dominant semantic modes. The semantic component therefore organizes candidate samples in a three-level concept graph and promotes coverage of underrepresented semantic regions while providing interpretable acquisition rationales. By balancing visual difficulty with semantic coverage, GSAL improves retrieval of subtle and rare targets that are often missed by uncertainty-only selection. Experiments on a proprietary thin-film defect, Pascal VOC and MS COCO dataset show consistent gains in label efficiency and rare-class retrieval over uncertainty-, diversity-, and hybrid-based baselines