Designing Effective Speech Interfaces

Toys and Education

Texas Instruments developed a children's toy called the Speak & Spell in the late 1970s. This toy would present spelling words to children by “sounding out” the word and waiting for the child to properly spell the word using keyboard input. It was an early and successful use of speech synthesis. The toy enjoyed high sales for over a decade without modification of the original design.

Toy manufacturers are researching numerous toys that have both speech recognition and speech synthesis capabilities. One factor inhibiting these applications is that most speech recognizers are calibrated for adult speakers. There is significant work ahead on recognition engines before these recognizers can handle children's speech sounds and patterns.

Most speech recognizers are calibrated for adult speakers.

In an education example, Swift Jewel publishes software called Talking Typing Tutor. The software is a CD-ROM-based tutorial that presents a series of keyboarding lessons. It uses speech synthesis to give users “real time” mentoring, such as telling them when to speed up or slow down during a lesson.