A Face Attention Technique for a Robot Able to Interpret Facial Expressions

Simplício, Carlos; Prado, Jose; Dias, Jorge

doi:10.1007/978-3-642-11628-5_36

Cited by 1 publication

(3 citation statements)

References 6 publications

(3 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For detecting features, still many of the current approaches do not automatically extract the features, do not consider time sequence frames, and it is common to divide the image in parts instead of analyzing the whole face image at once. A step forward was done in [18] and [13,10] were the features were detected automatically. About the classification techniques, we found on literature: template-based classification [25], fuzzy classification, ANN based classification [12], HMM based classification and also Bayesian classification [3] [17] and [18].…”

Section: Proposed Approaches To Automatic Emotionmentioning

confidence: 99%

“…Most state-of-the-art in the area of emotive robots do not run in real-time, being then still inapplicable to real cases of human-robot-interaction applications. However recent researches like [20] and [15] shows result about two Bayesian classifiers inside a structure for human-robot-interaction that are applicable to HRI in real-time. These classifiers have both the purpose to classify human emotional state among the scope {anger,fear,sad,neutral and happy}.…”

Section: Emotive Robotsmentioning

confidence: 99%

“…This structure is presented on figure 1. For the detection of features and classification of facial expressions we are using our previously proposed methods from [18]. For the classification of vocal expressions we are using a Bayesian model previously proposed by us in [15].…”

Section: Our Approachmentioning

confidence: 99%

“…After the BMM takes place, we have the detected emotion represented by variable HE, which stands for Human Emotion and vary among the scope {Anger, Fear, Sad, Happy, Neutral}. Next step is to compute the probability of response given the human emotion and the robotic SBP that Figure 1: Our system structure: Notice that two modalities signals are processed and then classified, Bayesian Classifier 1 was proposed in [15] and Bayesian Classifier 2 was proposed in [18]. Then, the fusion proposed in this paper takes place to reach to a fused detected emotion.…”

Section: Bayesian Mixture Modelmentioning

confidence: 99%

“…During the classification, as mentioned in [15] and [18], a likelihood table is filled out with the probabilities of leaf variables for a specific expression.…”

Section: Computational Synthesismentioning

confidence: 99%

See 4 more Smart Citations

Synthesis of Emotions on a <em>Human-Robot-Interactive</em> Platform

Prado¹,

Seneviratne²,

Dias³

2011

Biomechanics / 752: Robotics

View full text Add to dashboard Buy / Rent full text

This work explores our developments over a framework and over a platform for human-robot-interaction with emotions. The focus of our framework is on visuo-auditory perception and response. In other words, perception and response can be called analysis and synthesis; the analysis is responsible for the classification of human emotion and the synthesis is responsible for the synthetic expression that the robot must show. This paper is focused on the synthesis and also on how the synthesis can affect human engagement during an interactive conversation with the robot.

show abstract

Section: Proposed Approaches To Automatic Emotionmentioning

confidence: 99%

Section: Emotive Robotsmentioning

confidence: 99%

Section: Our Approachmentioning

confidence: 99%

Section: Bayesian Mixture Modelmentioning

confidence: 99%

“…During the classification, as mentioned in [15] and [18], a likelihood table is filled out with the probabilities of leaf variables for a specific expression.…”

Section: Computational Synthesismentioning

confidence: 99%

See 3 more Smart Citations