Semiotic prediction of driving behavior using unsupervised double articulation analyzer

Abstract

In this paper, we propose a novel semiotic prediction method for driving behavior based on double articulation structure. It has been reported that predicting driving behavior from its multivariate time series behavior data by using machine learning methods, e.g., hybrid dynamical system, hidden Markov model and Gaussian mixture model, is difficult because a driver's behavior is affected by various contextual information. To overcome this problem, we assume that contextual information has a double articulation structure and develop a novel semiotic prediction method by extending nonparametric Bayesian unsupervised morphological analyzer. Effectiveness of our prediction method was evaluated using synthetic data and real driving data. In these experiments, the proposed method achieved long-term prediction 2-6 times longer than some conventional methods.

Contextual scene segmentation of driving behavior based on double articulation analyzer

“…The effectiveness of the contextual scene segmentation, i.e., a symbolization framework of driving behavior based on physical and semantic perspectives has been reported on the basis that the symbolization enabled a long-term prediction of driving behavior [11]. In the present paper, we showed that the semantic symbolization is related to a human recognition of a driving scene.…”

Socio-Cyberphysical System for Parking Support

“…As a result, it is possible to build a conditional behavior model, which would let one know (to predict) how a human (e.g., a driver or a pedestrian) would act in a given situation. For example, the research of application of this technique to the driver behavior prediction has resulted in some positive results [26].…”

Towards prediction of driving behavior via basic pattern discovery with BP-AR-HMM

“…The methodology of Bayesian nonparametrics is one of the method of Bayesian statistics, attempting to learn the model complexity automatically according to training data [12,13]. Taniguchi et al utilized sHDP-HMM to model driving behaviors, and succeeded to segment driving time series [9]. In contrast, we utilized BP-AR-HMM that is an extension of AR-HMM as a Bayesian nonparametric approach.…”

“…Practically, some researchers have developed enthusiastically the indices of the risk of collision and the automatic emergency brake system for automotive vehicle, to suppress the number of traffic accidents [2,3]. And recently researchers turn to think about the estimation of driving scenes and the prediction of behaviors of drivers to realize novel driving support systems, in order to support drivers in diverse environment [4][5][6][7][8][9], not just to prevent collisions. If we can estimate driving scenes or driver's behaviors, it is possible to utilize the driving support system like the collision preventing system according to present driving scene, which is effective to prevent accidents beforehand.…”