OD-GCN: Object Detection Boosted by Knowledge GCN

TL;DR

This paper introduces OD-GCN, a novel object detection framework that integrates Graph Convolutional Networks to leverage high-level category relationships among objects, resulting in improved detection accuracy.

Contribution

The paper presents a new method combining GCN with existing detection models to utilize object relationships, enhancing detection performance.

Findings

Improved mAP by 1-5 percentage points on COCO dataset.

Enhanced detection confidence through high-level object relationships.

Visual analysis confirms the reasonableness of improvements.

Abstract

Classical CNN based object detection methods only extract the objects' image features, but do not consider the high-level relationship among objects in context. In this article, the graph convolutional networks (GCN) is integrated into the object detection framework to exploit the benefit of category relationship among objects, which is able to provide extra confidence for any pre-trained object detection model in our framework. In experiments, we test several popular base detection models on COCO dataset. The results show promising improvement on mAP by 1-5pp. In addition, visualized analysis reveals the benchmark improvement is quite reasonable in human's opinion.