Description:
Human-Object Interaction (HOI) detection is a crucial problem for comprehensive visual understanding. Many existing methods often exploit to integrate multiple level features into a framework to infer the interactions. However, most methods simply concatenate all these features that are not explicitly embedded in the feature learning for HOI detection. In this paper, we are trying to fuse these features explicitly using a multi-stream feature refinement network. The network extracts the visual features of humans, contexts, and objects, which receives the attentions from human poses, spatial configurations, and semantic prior knowledge of objects to refine these visual features, respectively. We verify our method on V-COCO and HICO-DET datasets with extensive experiments. The experimental results demonstrate that our method is a simple yet effective for HOI detection, achieving superior performance to those state-of-the-art methods