Incremental Learning of Statistical Motion Patterns With Growing Hidden Markov Models

Vasquez, Dizan; Fraichard, Thierry; Laugier, Christian

doi:10.1109/tits.2009.2020208

Cited by 87 publications

(24 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In other words, transitions between basic behaviors are only allowed if the corresponding nodes of a topological map over the feature space are connected. In the present work, the observation space is clustered incrementally using the instantaneous topological map (ITM) algorithm [54,89], which provides a discrete representation of the continuous feature space in the form of a graph where feature space regions are represented as nodes and adjacent regions are connected by edges. 3 The topological map is updated at every new observation O t in order to minimize the number of nodes as well as the model distorsion.…”

Section: Incremental Structure Learningmentioning

confidence: 99%

The intentional stance as structure learning: a computational perspective on mindreading

et al. 2015

View full text Add to dashboard Cite

Recent theories of mindreading explain the recognition of action, intention, and belief of other agents in terms of generative architectures that model the causal relations between observables (e.g., observed movements) and their hidden causes (e.g., action goals and beliefs). Two kinds of probabilistic generative schemes have been proposed in cognitive science and robotics that link to a "theory theory" and "simulation theory" of mindreading, respectively. The former compares perceived actions to optimal plans derived from rationality principles and conceptual theories of others' minds. The latter reuses one's own internal (inverse and forward) models for action execution to perform a look-ahead mental simulation of perceived actions. Both theories, however, leave one question unanswered: how are the generative models - including task structure and parameters - learned in the first place? We start from Dennett's "intentional stance" proposal and characterize it within generative theories of action and intention recognition. We propose that humans use an intentional stance as a learning bias that sidesteps the (hard) structure learning problem and bootstraps the acquisition of generative models for others' actions. The intentional stance corresponds to a candidate structure in the generative scheme, which encodes a simplified belief-desire folk psychology and a hierarchical intention-to-action organization of behavior. This simple structure can be used as a proxy for the "true" generative structure of others' actions and intentions and is continuously grown and refined - via state and parameter learning - during interactions. In turn - as our computational simulations show - this can help solve mindreading problems and bootstrap the acquisition of useful causal models of both one's own and others' goal-directed actions.

show abstract

Section: Incremental Structure Learningmentioning

confidence: 99%

The intentional stance as structure learning: a computational perspective on mindreading

et al. 2015

View full text Add to dashboard Cite

show abstract

“…A growing Hidden Markov Model (HMM) is proposed in [11] to maintain an online adaptive model that changes with incremental observations. Rather than Learning then Predicting, the patterns can be learned incrementally and in parallel with prediction.…”

Section: B Vehicle Behavior Analysismentioning

confidence: 99%

Vehicle motion intention reasoning using cooperative perception on urban road

Wei

Kim

Marczuk

et al. 2014

17th International IEEE Conference on Intelligent Transportation Systems (ITSC)

View full text Add to dashboard Cite

As autonomous vehicles are venturing on the urban road, reasoning other vehicles' motion intention is always essential for anticipatory decision-making. This paper considers a new class of motion intention reasoning problem using cooperative perception. Specifically, thanks to the extended perception range contributed by cooperative perception, we can explicitly consider the behavior correlations among multiple vehicles, where each vehicle's motion intention can be inferred using both the temporal observations and the other vehicles' motion intention expectation. As such, a Simplified variant of Coupled Hidden Markov Model (S-CHMM) is proposed for motion intention inference. The experiment on an autonomous vehicle shows that the proposed algorithm can improve the motion intention accuracy and provide efficient overtaking triggering for autonomous driving on urban road.

show abstract

“…Later, Vasquez et al [4] proposed an incremental model, so-called Growing Hidden Markov Models, to learn and predict the motion patterns.…”

Section: Introductionmentioning

confidence: 99%

“…Meanwhile, most of the mentioned works do not consider how the environment will affect the person movement (e.g. [1], [2], [3], [4], and [6]). In case of [5], it exploits the physical attribute information of the environment (such as building, car, and pavement), but it is basically used for separating the walkable and non-walkable area.…”

Section: Introductionmentioning

confidence: 99%

Human motion prediction considering environmental context

Ardiyanto

Miura

2015

2015 14th IAPR International Conference on Machine Vision Applications (MVA)

View full text Add to dashboard Cite

This paper describes an approach to predict the human motion. Instead of using a simple motion model as widely used, we take advantages of the environmental context, including the shape and structure, for predicting the human movement. First, we characterize the environment using a graph representation. Subsequently, we acquire the human trajectory tendency on each environment and build a probabilistic sequence model of the human motion. A particle filter-based predictor is then integrated into the system for generating possible future paths of the person. Evaluations on a real campus environment show the advantages of the proposed approach.

show abstract

Incremental Learning of Statistical Motion Patterns With Growing Hidden Markov Models

Cited by 87 publications

References 30 publications

The intentional stance as structure learning: a computational perspective on mindreading

The intentional stance as structure learning: a computational perspective on mindreading

Vehicle motion intention reasoning using cooperative perception on urban road

Human motion prediction considering environmental context

Contact Info

Product

Resources

About