Fig. 5From: Hidden Markov Model based stride segmentation on unsupervised free-living gait data in Parkinson’s disease patientsHMM structure of the transition sub-model: A left-right Markov chain from state \(s_{0}\) to \(s_{m}\), with transition probability \(p_{m,m+1}\), and self transition probability \(p_{m,m}\). Additionally, a transition from \(s_{m}\) to \(s_{0}\) was added. Sequences may start and end from any stateBack to article page