Speech Group Videos
The following videos show speech systems produced at various times the the Speech Group at Carnegie Mellon University. Many of them were produced before the era of video and were filmed in 16mm. The currently available versions were digitized. Due to the transfer, the quality of some videos is not that great.
They can be viewed using RealPlayer
software (.rm, .mpg) or Microsoft Media Player (.asf, .mpg).
- Hearsay [.mpg] (1976?) A very early recognition system created during the DARPA-SUR program. Hearsay is notable in that it took an AI approach to recognition; it introduced such cocepts as the Blackboard.
- Sphinx [.mpg] (1988) Sphinx was the first recognition system to combine speaker-independence, continuous speech and a large vocabulary. It also ran in near-real time on a general-purpose computer.
- Early CMU Speech Systems
[.mp4] (1988) Voice Calculator and Voice PID
- Examples of Voice Interfaces
[.mp4] (1989). A voice spreadsheet and an early version of the OM PID system.
- Spoken Language Interfaces: The OM System [.mp4] (1990)
The CMU Office Manager, a multi-application speech interface.
- Speech Recdognition in the Office Environment
[.mp4] (1991) Another view of the OM system.
- Speech recognition and understanding systems
[.rm] (July 1994) Two applications based on Sphinx, general dictation and the ATIS spoken language understanding system.
- The SpeechWear system [.rm] (June 1995)
- Information access through speech
[.rm] (1997) An early example of a mobile, browser-based system. It was the first system to incorporate ASR into the browser (Mosaic) and to implement html tags for speech.
- Communicator [.asf] (2000) An advanced mixed-initiative dialog system created during the DARPA Communicator program. It features the AGENDA dialog manager (later replaced as the Ravenclaw dialog manager).
- Universal Speech Interface [.mpg] (2000) The USI project developed a set of design principles that allows the same set of interaction to work across many different application. Much as current GUIs, USI provides a uniform experience across applications.
- Symphony [.avi] (2001)
A voice-enabled mobile maintenance system.
- Carnegie Mellon Team Talk
[.wmv] (2009) TeamTalk is a platform for multi-participant dialog between humans and robots.
Last modified: Wed Jun 26 19:09:40 EDT 2002