This chapter focus on the analysis of one special activity, Human-Object Interaction (HOI), which reflects the mutual relationship between human and objects in a specific application scenario. This activity is also a common behavior, in which the initiator generally has a clear purpose. The first step would be the detection of human-object interaction. Such a detection is to locate individuals who produce specific action behaviors and objects that interact with them in images or videos. By recognizing the action relationships between humans and objects, the human behavior can be analyzed and understand. The detection of human-object interaction is a task of locating humans and objects in a scene and predicting their interactions. This task can also be accomplished by identifying humans and objects with interactive behaviors. This chapter will introduce the principles of two-stage and single-stage methods for detecting human-object interaction relationships. Several typical detection methods for human-object interaction relationships with distinctive features are presented. This chapter will analyze the detection methods of human-object interaction relationship based on transformer and attention mechanism. In addition, some databases used for detecting human-object interaction relationships and several evaluation indicators for the performance evaluation of detection methods are provided.