Speech Perception

Speech perception research focuses on understanding how humans and machines process spoken language, aiming to improve both human communication in challenging acoustic environments and the accuracy of speech recognition systems. Current research utilizes large language models (LLMs) and deep neural networks (DNNs), such as Wav2Vec 2.0 and transformer-based architectures, to analyze acoustic features, contextual cues, and the interplay between auditory and visual information, investigating biases and limitations in these models. These advancements have implications for improving speech recognition technology, particularly for individuals with hearing impairments, and for developing more human-like and robust AI agents capable of natural communication.