Newest papers

Open-Vocabulary Models and Label Propagation for Uncertainty-aware Pseudo-Labeling of Object Detection Datasets

Jonathan Klees, Timo Heller, Matthias Rottmann · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.29788 · cited by 0 works (OpenAlex) · open-access PDF

Spanish

Los modelos de detección de potencial de objetos se basan en conjuntos de datos anotados de datos a gran escala que requieren un esfuerzo de anotación manual sustancial. Los costos de etiquetado aumentan aún más cuando se captura la incertidumbre de la anotación a través de etiquetas blandas, ya que se deben recopilar múltiples anotaciones para cada objeto. Sobre la base de estas representaciones, abordamos este desafío introduciendo un marco semi-supervisado para la anotación consciente de la incertidumbre de los conjuntos de datos de detección de objetos. Nuestro enfoque utiliza detectores de objetos de vocabulario abierto para generar propuestas regionales y representaciones de objetos semánticamente significativas. Sobre la base de estas representaciones, construimos un gráfico de similitud de objeto y propagamos información

Machine translation (Meta M2M100); the English original follows.

Abstract

Object detection models rely on large-scale annotated datasets that require substantial manual annotation effort. The labeling costs further increase when annotation uncertainty is captured through soft labels, as multiple annotations must be collected for each object. We address this challenge by introducing a semi-supervised framework for the uncertainty-aware annotation of object detection datasets. Our approach leverages open-vocabulary object detectors to generate region proposals and semantically meaningful object representations. Based on these representations, we construct an object-level similarity graph and propagate probabilistic label information. Our framework assigns soft labels to previously unlabeled objects, quantifying both aleatoric and epistemic uncertainty. We evaluate our method on four object detection benchmarks with soft-label annotations and demonstrate that high-quality annotations can be obtained with only a small fraction of manually labeled objects. Furthermore, we show that detectors trained on semi-automatically annotated data achieve performance comparable to models trained on fully manually annotated data. We achieve state-of-the-art downstream detection performance for the task of semi-supervised object detection on the COCO dataset, with 40.5% AP virtually matching the performance of the model trained on fully manually constructed data (40.9%) at only 1% of labeling costs. These results highlight the potential of open-vocabulary models and graph-based label propagation for efficient, accurate and uncertainty-aware pseudo-labeling.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Annotation, Computer science, Object detection, Artificial intelligence, Probabilistic logic, Object (grammar)

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange