A light-weight, data-driven segmentation method for multi-state Brownian trajectories
Ismail El Korde, Jason M. Lewis, Erik Clarkson, Tommy Dam, Peter Jönsson, Tobias Ambjörnsson, Joakim Stenhammar
This paper introduces a lightweight segmentation method for multi-state Brownian trajectories using Gaussian filtering and mixture modeling, showing high accuracy and lower computational load compared to deep learning or HMM approaches.
npj Biological Physics and Mechanics. · May 8, 2026Peer-reviewedBiophysics