English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent years have seen human robot collaboration (HRC) quickly emerged as ahot research area at the intersection of control, robotics, and psychology.While most of the existing work in HRC focused on either low-level human-awaremotion planning or HRC interface design, we are particularly interested in aformal design of HRC with respect to high-level complex missions, where it isof critical importance to obtain an accurate and meanwhile tractable humanmodel. Instead of assuming the human model is given, we ask whether it isreasonable to learn human models from observed perception data, such as thegesture, eye movements, head motions of the human in concern. As our initialstep, we adopt a partially observable Markov decision process (POMDP) model inthis work as mounting evidences have suggested Markovian properties of humanbehaviors from psychology studies. In addition, POMDP provides a generalmodeling framework for sequential decision making where states are hidden andactions have stochastic outcomes. Distinct from the majority of POMDP modellearning literature, we do not assume that the state, the transition structureor the bound of the number of states in POMDP model is given. Instead, we use aBayesian non-parametric learning approach to decide the potential human statesfrom data. Then we adopt an approach inspired by probably approximately correct(PAC) learning to obtain not only an estimation of the transition probabilitybut also a confidence interval associated to the estimation. Then, theperformance of applying the control policy derived from the estimated model isguaranteed to be sufficiently close to the true model. Finally, data collectedfrom a driver-assistance test-bed are used to train the model, whichillustrates the effectiveness of the proposed learning method.

POMDP Model Learning for Human Robot Collaboration