Mutual Supervision for Dense Object Detection

TL;DR

This paper introduces Mutual Supervision (MuSu), a novel training paradigm for dense object detectors where classification and regression heads supervise each other with mutually assigned samples, improving detection accuracy.

Contribution

The paper proposes a new mutual supervision framework that assigns training samples separately for classification and regression heads, enhancing their interaction and detector performance.

Findings

Mutual supervision guarantees convergence of trained detectors.

Significant improvements on MS COCO benchmark.

Tiling more anchors further enhances detection performance.

Abstract

The classification and regression head are both indispensable components to build up a dense object detector, which are usually supervised by the same training samples and thus expected to have consistency with each other for detecting objects accurately in the detection pipeline. In this paper, we break the convention of the same training samples for these two heads in dense detectors and explore a novel supervisory paradigm, termed as Mutual Supervision (MuSu), to respectively and mutually assign training samples for the classification and regression head to ensure this consistency. MuSu defines training samples for the regression head mainly based on classification predicting scores and in turn, defines samples for the classification head based on localization scores from the regression head. Experimental results show that the convergence of detectors trained by this mutual supervision is guaranteed and the effectiveness of the proposed method is verified on the challenging MS COCO benchmark. We also find that tiling more anchors at the same location benefits detectors and leads to further improvements under this training scheme. We hope this work can inspire further researches on the interaction of the classification and regression task in detection and the supervision paradigm for detectors, especially separately for these two heads.