Persian Speech

Research on Persian speech focuses on developing accurate and efficient automatic speech recognition (ASR) and emotion detection systems, addressing the challenges posed by a relatively low-resource language. Current efforts leverage transformer-based models and reservoir computing architectures, often incorporating techniques like data augmentation and imbalanced data handling to improve performance on tasks such as speech emotion recognition and lip reading, utilizing newly created large-scale datasets like Arman-AV and EmoPars. These advancements contribute to broader progress in multilingual speech processing and have practical applications in areas like customer service, opinion mining, and accessibility technologies.