Wearable-sensor activity analysis using semi-Markov models with a grammar
O Thomas, Peter Sunehag, G Dror, S Yun, S Kim, M Robards, A.J. Smola, D Green and P Saunders
Pervasive and Mobile Computing
Detailed monitoring of training sessions of elite athletes is an important component of their training. In this paper we describe an application that performs a precise segmentation and labeling of swimming sessions. This allows a comprehensive break-down of the training session, including lap times, detailed statistics of strokes, and turns. To this end we use semi-Markov models (SMM), a formalism for labeling and segmenting sequential data, trained in a max-margin setting. To reduce the computational complexity of the task and at the same time enforce sensible output, we introduce a grammar into the SMM framework. Using the trained model on test swimming sessions of diﬀerent swimmers provides highly accurate segmentation as well as perfect labeling of individual segments. The results are signiﬁcantly better that those achieved by discriminative Markov Models.