Detecting Stereotype Body Rocking Behavior Through Embodied Motion SensorsBy Krishna, Sreekar; Krishnan, Narayanan; Panchanathan, Sethuraman; RESNA Annual Conference - 2009,
Publication Date: 2011
Paper presents an assistive device for the detection of stereotypic body rocking behavior. Stereotypic behavior refers to any mannerism that is non functional and repetitive in nature, such as body rocking, head weaving, or repetitive vocal behaviors. Such behavior has been identified in individuals with varying physical and cognitive impairments, including blindness and autism, and can become a hindrance to social interactions and acceptance. The proposed detection device incorporates a triaxial accelerometer which is mounted on a hat and connects wirelessly to a PDA or cell phone that analyses the incoming motion patterns and detects the exact time instance of body rocking. An AdaBoost machine learning algorithm framework allows the system to understand and encode motion pattern differences between activities of everyday living and stereotypic body rocking. Training of the system was done offline on recorded data of rocking and non-rocking behavior collected with 10 participants who did not have any known stereotypic rocking behavior. For the collection of rocking data, participants used a rocking chair or sat on the ground while rocking, while non-rocking data were collected with participants performing everyday activities. To test the system’s ability to detect rocking, data where participants did various activities were interspersed randomly with rocking data. Testing showed that the system was able to detect rocking from non-rocking on an average of 95 percent and within one rock cycle. Future directions in the development of the device, including its use in self monitoring of individuals with stereotypic body rocking behavior, are discussed.
Published by: Rehabilitation Engineering & Assistive Technology Society of North America (RESNA) (Website:http://www.resna.org)
Rehabilitation Engineering & Assistive Technology Society of North America (RESNA) (Web Site: http://www.resna.org )
Link to text: http://web.resna.org/conference/proceedings/2009/Other/Student%20Papers/KrishnaS.html