Box-Level Class-Balanced Sampling for Active Object Detection

TL;DR

This paper introduces a class-balanced sampling strategy for active object detection that improves class balance and pseudo label accuracy, leading to state-of-the-art results on benchmark datasets.

Contribution

It proposes a novel class-balanced sampling method and a task-aware pseudo labelling strategy for box-level active learning in object detection.

Findings

Achieves state-of-the-art performance on benchmark datasets.

Effectively balances classes during active learning, improving detection accuracy.

Enhances pseudo label quality with task-aware strategies.

Abstract

Training deep object detectors demands expensive bounding box annotation. Active learning (AL) is a promising technique to alleviate the annotation burden. Performing AL at box-level for object detection, i.e., selecting the most informative boxes to label and supplementing the sparsely-labelled image with pseudo labels, has been shown to be more cost-effective than selecting and labelling the entire image. In box-level AL for object detection, we observe that models at early stage can only perform well on majority classes, making the pseudo labels severely class-imbalanced. We propose a class-balanced sampling strategy to select more objects from minority classes for labelling, so as to make the final training data, \ie, ground truth labels obtained by AL and pseudo labels, more class-balanced to train a better model. We also propose a task-aware soft pseudo labelling strategy to increase the accuracy of pseudo labels. We evaluate our method on public benchmarking datasets and show that our method achieves state-of-the-art performance.