| recognition and voice recognition are both rapidly | | | | While speech recognition is the process of |
| progressing technologies with numerous | | | | converting speech to digital data, voice recognition |
| applications that can enhance convenience, | | | | is aimed toward identifying the person who is |
| security, law enforcement efforts, and more. | | | | speaking. |
| Though speech recognition and voice | | | | Voice recognition works by analyzing the features |
| recognition are often used interchangeably, | | | | of speech that differ between individuals. |
| they are different technologies with radically | | | | Everyone has a unique pattern of speech |
| different objectives. | | | | stemming from their anatomy (the size and |
| What is speech recognition? | | | | shape of the mouth and throat) and behavioral |
| Speech recognition is the process of capturing | | | | patterns (their voice’s pitch, their speaking |
| spoken words using a microphone or telephone | | | | style, accent, and so on). |
| and converting them into a digitally stored set of | | | | The applications of voice recognition are markedly |
| words. | | | | different from those of speech recognition. Most |
| The quality of a speech recognition systems are | | | | commonly, voice recognition technology is used to |
| assessed according to two factors: its accuracy | | | | verify a speaker’s identity or determine an |
| (error rate in converting spoken words to digital | | | | unknown speaker’s identity. Speaker |
| data) and speed (how well the software can keep | | | | verification and speaker identification are both |
| up with a human speaker). | | | | common types of voice recognition. |
| Speech recognition technology has endless | | | | Speaker verification is the process of using a |
| applications. Commonly, such software is used for | | | | person’s voice to verify that they are |
| automatic translations, dictation, hands-free | | | | who they say they are. Essentially, a |
| computing, medical transcription, robotics, | | | | person’s voice is used like a fingerprint. |
| automated customer service, and much more. If | | | | Once a sample of their speech is recorded, a |
| you have ever paid a bill over the phone using an | | | | person’s speech patterns are tested |
| automated system, you have probably benefited | | | | against a database to see if their voice matches |
| from speech recognition software. | | | | their claimed identity. |
| Speech recognition technology has made huge | | | | Most commonly, speaker verification is applied to |
| strides within the last decade. However, speech | | | | situations where secure access is needed. Such |
| recognition has its weaknesses and nagging | | | | systems operate with the user’s |
| problems. Current technology is a long way away | | | | knowledge and cooperation. |
| from recognizing conversational speech, for | | | | Speaker identification is the process of |
| example. | | | | determining an unknown speaker’s identity. |
| Despite its shortcomings, speech recognition is | | | | Unlike speaker verification, speaker identification is |
| quickly growing in popularity. Within the next few | | | | usually covert and done without the user’s |
| years, experts say that speech recognition will be | | | | knowledge. |
| the norm in phone networks the world over. Its | | | | For example, speaker identification can be used to |
| spread will be aided by the fact that voice is the | | | | identify a criminal solely by their voice. In this |
| only option for controlling automated services in | | | | situation, a sample of their voice would be |
| places where touch tone phones are uncommon. | | | | checked against a database of criminals’ |
| What is voice recognition? | | | | voices until a match is found. |