Loading...

Concept Extraction of Sequential Patterns for Imitative Learning

Arjomand Aghaee, Ehsan | 2015

556 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: English
  3. Document No: 47759 (55)
  4. University: Sharif University of Technology, International Campus, Kish Island
  5. Department: Science and Engineering
  6. Advisor(s): Bagheri Shouraki, Saeed
  7. Abstract:
  8. The aim of this thesis is the concept extraction of sequential patterns for imitative learning for humanoid robots. In such a way that an existent that has the physical and cognitive similarities, begins to extract concepts and learns by observing the behavior of the other existent. In this project, it is assumed a humanoid robot that can understand the concepts such as hello, goodbye and different concepts and does the corresponding actions from the visual and auditory information. In this thesis, a new model has been presented to eliminate the improper and meaningless elasticity in patterns sequence, such as changes in accent or elasticity in movements. This model is called the fuzzy elastic matching machine (FEMM). Its structural is a little distinct from common pattern recognition methods such as hidden Markov model or neural networks. The main foundations of the calculations is on the fuzzy logic. It has caused this model to have more resistant against environmental changes. It used the adaptive network fuzzy inference system which is the combination of neural networks and takagi -Sugeno fuzzy modeling in the training and modeling section. In this algorithm, the audio and visual observations data are clustered sequentially by a human for separate FEMMs. Clusters should be meaningful in the human points of view. After clustering, the classified information goes to the nodes of the related model and the fuzzy model is built by them. Simultaneously, the time information on any node is modeled as fuzzy parameter as well to allow the system to be taught to stand to the elasticity of the time. In the test phase, input audio and visual patterns are compared with the made fuzzy rules base on the training phase simultaneously. And finally a pattern will be recognized according to category, speech and relative motion
  9. Keywords:
  10. Fuzzification ; Pattern Recognition ; Elasticity ; Fuzzy Elastic Matching Machine ; Speech Recognition ; Fuzzy Modeling ; Mixture Model ; Adaptive-Network-Based Fuzzy Inference System

 Digital Object List

 Bookmark

No TOC