iCAN: Instance-Centric Attention Network for Human-Object Interaction Detection

TL;DR

This paper introduces iCAN, an instance-centric attention network that improves human-object interaction detection by dynamically focusing on relevant image regions based on individual instance appearances, leading to better recognition accuracy.

Contribution

The paper proposes a novel attention module that leverages instance appearance cues to enhance HOI detection, outperforming existing methods.

Findings

Outperforms state-of-the-art on COCO and HICO-DET datasets.

Effectively highlights relevant image regions for interaction recognition.

Demonstrates the importance of instance-centric attention in HOI tasks.

Abstract

Recent years have witnessed rapid progress in detecting and recognizing individual object instances. To understand the situation in a scene, however, computers need to recognize how humans interact with surrounding objects. In this paper, we tackle the challenging task of detecting human-object interactions (HOI). Our core idea is that the appearance of a person or an object instance contains informative cues on which relevant parts of an image to attend to for facilitating interaction prediction. To exploit these cues, we propose an instance-centric attention module that learns to dynamically highlight regions in an image conditioned on the appearance of each instance. Such an attention-based network allows us to selectively aggregate features relevant for recognizing HOIs. We validate the efficacy of the proposed network on the Verb in COCO and HICO-DET datasets and show that our approach compares favorably with the state-of-the-arts.