Label-Efficient Object Detection via Region Proposal Network Pre-Training

TL;DR

This paper introduces a pre-training method for the region proposal network (RPN) in object detection, significantly enhancing detection performance, especially in label-scarce scenarios, by reducing localization errors.

Contribution

It proposes a novel pretext task for RPN pre-training, addressing the gap of detection-specific module pre-training and improving downstream detection and segmentation tasks.

Findings

Consistent performance improvements over non-pre-trained RPNs.

Largest gains observed in label-scarce settings.

Effective pre-training reduces localization errors.

Abstract

Self-supervised pre-training, based on the pretext task of instance discrimination, has fueled the recent advance in label-efficient object detection. However, existing studies focus on pre-training only a feature extractor network to learn transferable representations for downstream detection tasks. This leads to the necessity of training multiple detection-specific modules from scratch in the fine-tuning phase. We argue that the region proposal network (RPN), a common detection-specific module, can additionally be pre-trained towards reducing the localization error of multi-stage detectors. In this work, we propose a simple pretext task that provides an effective pre-training for the RPN, towards efficiently improving downstream object detection performance. We evaluate the efficacy of our approach on benchmark object detection tasks and additional downstream tasks, including instance segmentation and few-shot detection. In comparison with multi-stage detectors without RPN pre-training, our approach is able to consistently improve downstream task performance, with largest gains found in label-scarce settings.