DESIGN AND DEVELOPMENT OF FUSION FRAMEWORK USING PHONEMES AND MORPHEMES FOR SPOKEN WORD RECOGNITION

Abstract

Spoken word recognition involves identifying words from spoken input. It specifically centers on recognizing and comprehending individual words within spoken language. Phonemes and morphemes play crucial roles in spoken word recognition. Phonemes are the smallest units of sound in a language, and they help differentiate between different words based on their pronunciation. Morphemes, on the other hand, are the smallest units of meaning in language, such as prefixes, suffixes, and root words. In spoken word recognition, phonemes help in distinguishing between words that sound similar but have different meanings. Morphemes provide additional context and meaning to words, aiding in understanding the overall message conveyed by the spoken language. By analyzing phonemes and morphemes, speech recognition systems can accurately identify and understand spoken words, enhancing their ability to convert spoken language into written text or commands. The study of spoken word recognition spans various fields like phonetics, linguistics, psychology, cognitive science, psycholinguistics, and computer science. Recent progress in deep learning and pre-trained models has transformed this field. These advancements allow for combining phonological and morphological parsing techniques, boosting the precision and effectiveness of recognizing words from spoken input. Uniting speech and understanding demands collaboration across multiple disciplines, reflecting the intricate and captivating nature of this research domain. This thesis contributes to the advancement of spoken word recognition technology by providing a nuanced understanding of phonological and morphological features and offering a versatile fusion framework. The proposed system has potential applications in various fields, including speech processing, natural language understanding, and human-computer interaction. This thesis focuses on the design and development of a comprehensive fusion framework to improve spoken word recognition by integrating phonemes and morphemes.