Published Paper Details:

Speech and speaker recognition technologies are one of the emerging technologies in today's world that have obtained rapid advancements. They're becoming integral components in various applications such as virtual assistants, authentication systems, and customer service platforms. The primary purpose of speech recognition is to transcribe spoken language into text and that of speaker recognition is to identify or verify an individual's identity based on vocal characteristics. This paper presents an exhaustive survey of techniques both in this and other fields which include acoustic-phonetic approach, pattern recognition, and artificial intelligence-based models to develop hybrid systems that integrate all the methods for improved accuracy. Also, this paper explores various techniques of speaker recognition. Here, again, one technique considered is the feature extraction method that is necessary to distinguish speakers from one another. Despite these innovations, however, current systems of speech recognition face problems like variability in accents, noisy environments, and data insufficient for training models to lead to deterioration in model performance. Complexity in complexities of model training further hampers progress toward the development of highly accurate adaptive systems for recognition. In discussing these challenges and surveying the current landscape of speech and speaker recognition technologies, insights are provided not only into their potential but also further developments needed for applications over wider spectra and more effectively in the real world.