Types of Voice Recognition Software

Voice recognition software offers us the science fiction experience of actually talking to our computers, and the system responding to what is said. Available technology is not yet to the level of the infamous HAL 9000 from the film 2001, but voice recognition software is being used in many hands-free applications.

VR vs SR

The term "Voice Recognition" is really a misnomer. What most people mean when using that term is "Speech Recognition," which brings us to a branch of software.

Voice Recognition

The term Voice Recognition, used correctly, might be better understood if it were termed Voice Security. In this sense, software is designed to recognize a specific voice and respond only to that user.

Speech Recognition

Speech Recognition means the computer hears spoken words and translates them into commands or text. The software may be programmed to respond to only one or many users.

Speech Recognition Types

Speech Recognition software branches out into three types: Discrete word recognition; connected word recognition; and continuous speech. Continuous systems respond to human normal speech; they are difficult to implement, and few exist.

Discrete Word

Discrete Word systems generally respond to many users, but recognize only specific, well-enunciated single words.

Connected Word

Commercial packages for medical, legal and business professional transcription use connected word packages in which a specific cadence and speaking voice must be used after training the software.

Resources

• Voice Recognition