What question did this study set out to answer?

The aim is to enhance active learning methods for object detection by reducing computational overhead and improving efficiency.

March 7, 2026Open Access

Uncertainty3D: A Lightweight Tri-Dimensional Uncertainty Framework for CNN-Based Active Learning in Object Detection

Key Points

The aim is to enhance active learning methods for object detection by reducing computational overhead and improving efficiency.
Introduced Uncertainty3D, a lightweight uncertainty proxy for CNN-based detectors.
Developed a tri-dimensional formulation capturing inconsistencies in position, scale, and category.
Evaluated on PASCAL VOC and MS COCO datasets using Faster R-CNN and RetinaNet.
Achieved competitive mean Average Precision (mAP) compared to representative baselines.
Estimated uncertainty approximately 3–4 times faster than augmentation-based methods.

Abstract

In object detection, annotation cost and computational efficiency are important factors in iterative model improvement under standard benchmark settings. Active learning (AL) addresses this challenge by selecting informative samples for labeling; however, many detection-oriented AL methods incur substantial overhead due to repeated inference (e.g., augmentation-based consistency). This paper introduces Uncertainty3D, a lightweight uncertainty proxy designed for standard CNN-based object detectors. It leverages native pre-NMS predictions to estimate sample informativeness using a single forward pass. We propose a tri-dimensional formulation that captures inconsistencies in position, scale, and category across proposal-consistent predictions. Experiments on PASCAL VOC and MS COCO using representative CNN-based detectors (Faster R-CNN and RetinaNet) show competitive mAP versus representative baselines and about 3–4× faster uncertainty estimation than augmentation-based baselines.

Read Full Paperexternally

Perguntar à IA

Bookmark

View Full Paper