Anchor-Intermediate Detector: Decoupling and Coupling Bounding Boxes for Accurate Object Detection

TL;DR

This paper introduces the Anchor-Intermediate Detector (AID) that decouples and couples bounding box corners to improve object detection accuracy, outperforming baseline models on MS COCO without additional bells and whistles.

Contribution

The paper proposes the Box Decouple-Couple (BDC) strategy and a novel AID model with dual heads, enhancing bounding box prediction accuracy in object detection.

Findings

AID outperforms RetinaNet and GFL by 2.4 and 1.2 AP on MS COCO.

BDC strategy effectively decouples and couples bounding box corners.

The corner-aware head improves corner scoring for better detection.

Abstract

Anchor-based detectors have been continuously developed for object detection. However, the individual anchor box makes it difficult to predict the boundary's offset accurately. Instead of taking each bounding box as a closed individual, we consider using multiple boxes together to get prediction boxes. To this end, this paper proposes the \textbf{Box Decouple-Couple(BDC) strategy} in the inference, which no longer discards the overlapping boxes, but decouples the corner points of these boxes. Then, according to each corner's score, we couple the corner points to select the most accurate corner pairs. To meet the BDC strategy, a simple but novel model is designed named the \textbf{Anchor-Intermediate Detector(AID)}, which contains two head networks, i.e., an anchor-based head and an anchor-free \textbf{Corner-aware head}. The corner-aware head is able to score the corners of each bounding box to facilitate the coupling between corner points. Extensive experiments on MS COCO show that the proposed anchor-intermediate detector respectively outperforms their baseline RetinaNet and GFL method by $\sim$2.4 and $\sim$1.2 AP on the MS COCO test-dev dataset without any bells and whistles. Code is available at: https://github.com/YilongLv/AID.