Implicit human intention inference through gaze cues for people with limited motion ability

The promising assistive technologies bring the hope that enlightens the independent daily living for the elderly and disabled people. However, most modern human-machine communication means is not affordable to those people with very limited motion ability to effectively express their service requests. In the paper, we presented a novel interaction framework which can facilitate the communication between human and assistive devices. In the framework, human intention is inferred implicitly by monitoring the gaze movements. The advantage of this framework is that gaze-based communication requires very little effort from the user and most elderly and disabled people with motion impairment retain the visual capability. The architecture of the presented framework and its effectiveness were introduced and validated. The relationship between human intentions with gaze behaviors was further discussed. This work is expected to simplify the human-machine interaction, consequently enhancing the adoption of assistive technologies and the user's independence in daily living.

I. Introduction

Human's motion ability got impaired and even severely lost due to aging and critical injury or diseases such as spinal traumas, paralysis, stroke, amputation, and Parkinson's. For those people, it will be difficult or impossible to independently carry out normal activities in daily living, like standing-up, dressing, cooking, eating etc.