The MobiFall Dataset

• DOI: 10.4018/ijmstr.2014010103
• Publication Date: 2014-10-10T19:21:25Z
• Authors: George Vavoulas; Matthew Pediaditis; Charikleia Chatzaki; Emmanouil G. Span...; Manolis Tsiknakis
• Keywords: Activity Recognition

Fall detection is receiving significant attention in the field of preventive medicine, wellness management and assisted living, especially for elderly. As a result, several fall systems are reported research literature or exist as commercial systems. Most them use accelerometers and/ gyroscopes attached on person's body primary signal sources. These either discrete sensors part product designed specifically this task that embedded mobile devices such smartphones. The latter approach has advantage offering well tested widely available communication services, e.g. calling emergency when necessary. Nevertheless, automatic continues to present challenges, with recognition type being most critical. aim work introduce human activity dataset be used testing new methods, performing objective comparisons between different algorithms recognition, based inertial-sensor data from contains signals recorded accelerometer gyroscope latest technology smartphone four types falls nine activities daily living. Utilizing dataset, results an elaborate evaluation machine learning-based classification presented discussed detail.