Box-Level Active Detection

TL;DR

This paper introduces a box-level active detection framework that efficiently selects informative object boxes for annotation, reducing labeling effort while maintaining high detection performance, and outperforms existing methods on standard datasets.

Contribution

It proposes a novel box-level active detection framework with a complementary pseudo active strategy, improving annotation efficiency and detection accuracy over prior image-level methods.

Findings

Achieves 100% VOC0712 performance with only 19% annotations.

Yields up to 4.3% mAP improvement on COCO over competitors.

Surpasses 90% COCO performance with 85% label reduction.

Abstract

Active learning selects informative samples for annotation within budget, which has proven efficient recently on object detection. However, the widely used active detection benchmarks conduct image-level evaluation, which is unrealistic in human workload estimation and biased towards crowded images. Furthermore, existing methods still perform image-level annotation, but equally scoring all targets within the same image incurs waste of budget and redundant labels. Having revealed above problems and limitations, we introduce a box-level active detection framework that controls a box-based budget per cycle, prioritizes informative targets and avoids redundancy for fair comparison and efficient application. Under the proposed box-level setting, we devise a novel pipeline, namely Complementary Pseudo Active Strategy (ComPAS). It exploits both human annotations and the model intelligence in a complementary fashion: an efficient input-end committee queries labels for informative objects only; meantime well-learned targets are identified by the model and compensated with pseudo-labels. ComPAS consistently outperforms 10 competitors under 4 settings in a unified codebase. With supervision from labeled data only, it achieves 100% supervised performance of VOC0712 with merely 19% box annotations. On the COCO dataset, it yields up to 4.3% mAP improvement over the second-best method. ComPAS also supports training with the unlabeled pool, where it surpasses 90% COCO supervised performance with 85% label reduction. Our source code is publicly available at https://github.com/lyumengyao/blad.