Automatic Speech Recognition (ASR) systems translate spoken utterances into text. ASRs are a crucial component of conversational assistants like Apple’s Siri, Amazon’s Alexa, etc. They are also used for closed captioning, interfacing with automated cars or homes, voice-biometrics, and provide improved access to technology. The global market for services enabled by ASR engines is estimated to grow to $20.9 billion by 2026.
In this talk, Nikhita Sharma will provide an overview of how ASR systems work and discuss the challenges associated with training and using such systems. She will also touch on the history of ASR systems, their applications, and limitations.
Nikhita Sharma is a Computer Scientist at Vail Systems, Chicago. She has over four years of experience in machine learning research, primarily focused on conversational AI, Automatic Speech Recognition (ASR), software engineering and model deployment. She has a master’s in computer science from Iowa State University.