Face-Computer Interface (FCI): Intent Recognition Based on Facial Electromyography (fEMG) and Online Human-Computer Interface With Audiovisual Feedback

• DOI: 10.3389/fnbot.2021.692562
• Publication Date: 2021-07-16T11:33:57Z
• Authors: Bo Zhu; Daohui Zhang; Yaqi Chu; Xingang Zhao; Lixin Zhang; Lina Zhao
• Keywords: Interface (matter); Facial muscles; Facial electromyography; Feature (linguistics)

Patients who have lost limb control ability, such as upper amputation and high paraplegia, are usually unable to take care of themselves. Establishing a natural, stable, comfortable human-computer interface (HCI) for controlling rehabilitation assistance robots other controllable equipments will solve lot their troubles. In this study, complete limbs-free face-computer (FCI) framework based on facial electromyography (fEMG) including offline analysis online mechanical was proposed. Six movements related eyebrows, eyes, mouth were used in FCI. the stage, 12 models, eight types features, three different feature combination methods model inputing studied compared detail. four well-designed sessions introduced robotic arm drinking water task ways (by touch screen, by fEMG with without audio feedback) verification performance comparison proposed FCI framework. Three features one an average recognition accuracy 95.3%, maximum 98.8%, minimum 91.4% selected use scenarios. contrast, way feedback performed better than that feedback. All subjects completed few minutes The smallest time difference between screen under only 1.24 0.37 min, respectively.