Point-to-Box Network for Accurate Object Detection via Single Point Supervision

TL;DR

This paper introduces P2BNet, a novel point-to-box network that constructs high-quality proposal bags for single point supervised object detection, significantly narrowing the performance gap with bounding box supervision.

Contribution

P2BNet provides a lightweight, anchor-like proposal generation method and an instance-level bag construction, improving detection accuracy in point supervision tasks.

Findings

Over 50% improvement in mean average precision on MS COCO

Effective proposal generation and bag construction strategies

Bridging the performance gap between point and bounding box supervision

Abstract

Object detection using single point supervision has received increasing attention over the years. However, the performance gap between point supervised object detection (PSOD) and bounding box supervised detection remains large. In this paper, we attribute such a large performance gap to the failure of generating high-quality proposal bags which are crucial for multiple instance learning (MIL). To address this problem, we introduce a lightweight alternative to the off-the-shelf proposal (OTSP) method and thereby create the Point-to-Box Network (P2BNet), which can construct an inter-objects balanced proposal bag by generating proposals in an anchor-like way. By fully investigating the accurate position information, P2BNet further constructs an instance-level bag, avoiding the mixture of multiple objects. Finally, a coarse-to-fine policy in a cascade fashion is utilized to improve the IoU between proposals and ground-truth (GT). Benefiting from these strategies, P2BNet is able to produce high-quality instance-level bags for object detection. P2BNet improves the mean average precision (AP) by more than 50% relative to the previous best PSOD method on the MS COCO dataset. It also demonstrates the great potential to bridge the performance gap between point supervised and bounding-box supervised detectors. The code will be released at github.com/ucas-vg/P2BNet.