The Speech Technologies group focuses on voice technologies and their use for advanced services and applications. We create technology components, solutions, frameworks, and services that enhance the experience and capabilities offered to mobile users and enterprises. The group's expertise covers a wide spectrum of technologies for speech analysis and classification, speech synthesis, speaker identification and voice biometrics, speech recognition, spoken information retrieval, and web services and applications.
Currently, the group's activities focus on three areas:
- Text-to-speech - We develop state-of-the-art expressive text-to-speech technology for delivering information and interacting with enterprise customers. The technology is currently offered as a service in the IBM Watson Developer Cloud.
- Voice biometrics - With the rapid growth of the mobile Internet and smart phones, security shortcomings in the mobile environment have shifted the focus to strong authentication that can be easily used by mobile users. We develop advanced voice biometrics technology for mobile security solutions, particularly in multifactor biometric authentication, applicable across industries.
- Speech analytics – We develop technologies and solutions for analysis and mining of spoken data, including emotion detection, audio classification and Affective Computing/segmentation . These technologies can be used for analytics of spoken data as well as for affect aware human computer interaction.