CBR: Fuzzified Case Retrieval Approach for Facial Expression Recognition

A. Khanum, M. Zubair Shafiq, and E. Muhammad (Pakistan)

Keywords

Casebased Reasoning, Facial Expression Recognition, Fuzzy Rule Based System, HumanComputer Interaction, Intelligent System Design

Abstract

We present a new scheme for the recognition of facial expressions from a set of facial features using fuzzy enhanced Case-based Reasoning (CBR). Facial expression recognition has become the cornerstone of human-computer interaction (HCI) systems. It has wide range applications information systems and e-commerce such as intelligent desktops and intelligent web agents. Our system is the integration of two fundamental paradigms of AI, i.e., Fuzzy logic and Case Based Reasoning. Fuzzy logic is embedded into our CBR system for improved case retrieval. A dual, fuzzy similarity determination module is the core of our system and works within the Case-based Reasoning system. The advantages from Fuzzy Rule Based Systems (FRBS) like linguistic modeling and fault tolerance when combined with the reasoning capability of CBR greatly improved the sophistication and utility of our system. A reinforcing combination of two approaches might lead to exciting applications that are yet to be envisaged.

Important Links:



Go Back