scholarly journals Vision-Based Attentiveness Determination Using Scalable HMM Based on Relevance Theory

Sensors ◽  
2019 ◽  
Vol 19 (23) ◽  
pp. 5331 ◽  
Author(s):  
Prasertsak Tiawongsombat ◽  
Mun-Ho Jeong ◽  
Alongkorn Pirayawaraporn ◽  
Joong-Jae Lee ◽  
Joo-Seop Yun
Keyword(s):  
Markov Model ◽  
Detection Rate ◽  
State Transition ◽  
Transition Probabilities ◽  
Hidden Markov ◽  
Relevance Theory ◽  
Image Sequences ◽  
Human Robot Interaction ◽  
Robot Interaction ◽  
Recent Attention

Attention capability is an essential component of human–robot interaction. Several robot attention models have been proposed which aim to enable a robot to identify the attentiveness of the humans with which it communicates and gives them its attention accordingly. However, previous proposed models are often susceptible to noisy observations and result in the robot’s frequent and undesired shifts in attention. Furthermore, most approaches have difficulty adapting to change in the number of participants. To address these limitations, a novel attentiveness determination algorithm is proposed for determining the most attentive person, as well as prioritizing people based on attentiveness. The proposed algorithm, which is based on relevance theory, is named the Scalable Hidden Markov Model (Scalable HMM). The Scalable HMM allows effective computation and contributes an adaptation approach for human attentiveness; unlike conventional HMMs, Scalable HMM has a scalable number of states and observations and online adaptability for state transition probabilities, in terms of changes in the current number of states, i.e., the number of participants in a robot’s view. The proposed approach was successfully tested on image sequences (7567 frames) of individuals exhibiting a variety of actions (speaking, walking, turning head, and entering or leaving a robot’s view). From these experimental results, Scalable HMM showed a detection rate of 76% in determining the most attentive person and over 75% in prioritizing people’s attention with variation in the number of participants. Compared to recent attention approaches, Scalable HMM’s performance in people attention prioritization presents an approximately 20% improvement.

A model based on hidden graphic evaluation and review technique network to evaluate reliability and lifetime of multi-state systems

2018 ◽  
Vol 233 (3) ◽  
pp. 369-378 ◽  
Author(s):  
Wenjie Dong ◽  
Sifeng Liu ◽  
Zhigeng Fang ◽  
Yingsai Cao ◽  
Ye Ding
Keyword(s):  
Markov Model ◽  
Hidden Markov Model ◽  
Network Model ◽  
State Transition ◽  
Likelihood Function ◽  
Hidden Markov ◽  
Reliability Assessment ◽  
Assessment Method ◽  
Performance Degradation ◽  
State System

The essence of multi-state system performance degradation is a process of deteriorating state transition. On the basis of hidden Markov model and graphic evaluation and review technique network, this article proposes a new reliability assessment method called hidden graphic evaluation and review technique network model for multi-state system. Specifically, nodes in graphic evaluation and review technique network represent hidden states of a system at different deteriorating times, and they can be expanded through a series of observable sequences. Baum–Welch algorithm in hidden Markov model is introduced to train parameters and when logarithmic likelihood function of the output reaches convergent, we can estimate the most probable output state and obtain the state transition probability eventually. Suppose performance degradation amount between different nodes follows gamma distribution, a method of improved maximum likelihood function is introduced to estimate parameters. According to signal flow graph theory and Mason formula, equivalent transfer function from the initial node to any other nodes can be obtained, then expectation and variance of performance degradation amount can be presented. In the real case study, we compare the reliability assessment method proposed in this article with other two traditional methods, which show the rationality of hidden graphic evaluation and review technique network model.

scholarly journals A Decision Analytical Model Investigating Cost-Effectiveness of Erlotinib

2020 ◽  
Author(s):  
Mohammed Alam
Keyword(s):  
Cost Effectiveness ◽  
Markov Model ◽  
Analytical Model ◽  
State Transition ◽  
Transition Probabilities ◽  
Qaly Gain ◽  
Health State ◽  
Time Varying ◽  
Decision Analytical Model ◽  
The Cost

Background: A decision analytical model investigating cost-effectiveness of Erlotinib was submitted to the UK NICE (National Institute for Health and Care Excellence), which was not based on actual health-state transition probabilities, leading to structural uncertainty in the model. The study adopted a Markov state-transition model for investigating the cost-effectiveness of Erlotinib versus Best Supportive Care (BSC) as a maintenance therapy for patients with non-small cell lung cancer (NSCLC). Methods: Unlike manufacturer submission (MS), the Markov model was governed by transition probabilities, and allowed a negative post-progression survival (PPS) estimate to appear in later cycle. Using published summary survival data, the study employs three fixed- and time-varying approaches to estimate state transition probabilities that are used in a restructured model. Results: Post-progression probabilities and probabilities of death for Erlotinib were different than fixed-transition approaches. The best fitting curves are achieved for both PPS and probability of death across the time for which data were available, but the curves start diverging towards the end of this period. The Markov model which extrapolates the curves forward in time suggests that this difference between a time-varying and fixed-transition becomes even greater. Our models produce an ICER of £54k -£66k per QALY gain, which is comparable to an ICER presented in the MS (£55k/QALY gain). Conclusions: Results from restructured Markov models show robust cost-effectiveness results for Erlotinib vs BSC. Although these are comparable to manufacturer submissions, in terms of magnitude, they vary, and which are crucial for interventions falling near a threshold value. The study will further explore the cost-effectiveness of therapies for NSCLC in Qatar.

scholarly journals State-Transition Modeling of Human–Robot Interaction for Easy Crowdsourced Robot Control

Sensors ◽  
2020 ◽  
Vol 20 (22) ◽  
pp. 6529
Author(s):  
Masaya Iwasaki ◽  
Mizuki Ikeda ◽  
Tatsuyuki Kawamura ◽  
Hideyuki Nakanishi
Keyword(s):  
User Interface ◽  
Social Presence ◽  
Robot Control ◽  
State Transition ◽  
Human Robot Interaction ◽  
Sales Promotion ◽  
Robot Interaction ◽  
Sensor Information ◽  
Transition Modeling ◽  
First Time

Robotic salespeople are often ignored by people due to their weak social presence, and thus have difficulty facilitating sales autonomously. However, for robots that are remotely controlled by humans, there is a need for experienced and trained operators. In this paper, we suggest crowdsourcing to allow general users on the internet to operate a robot remotely and facilitate customers’ purchasing activities while flexibly responding to various situations through a user interface. To implement this system, we examined how our developed remote interface can improve a robot’s social presence while being controlled by a human operator, including first-time users. Therefore, we investigated the typical flow of a customer–robot interaction that was effective for sales promotion, and modeled it as a state transition with automatic functions by accessing the robot’s sensor information. Furthermore, we created a user interface based on the model and examined whether it was effective in a real environment. Finally, we conducted experiments to examine whether the user interface could be operated by an amateur user and enhance the robot’s social presence. The results revealed that our model was able to improve the robot’s social presence and facilitate customers’ purchasing activity even when the operator was a first-time user.

Sign in / Sign up

Export Citation Format

Share Document