Speech recognition, Voice Recognition, is a method of speech analysis in which a computer-based system with automatic speech recognition analyzes, classifies and stores the input speech information. The stored speech information is related to the spoken words, to their meaning and in relation to the characteristic features of the speaker.
The different approaches to speech recognition are application specific. It can stand for speech input in text entry, for machine and automation control, and for security-related functions with biometric data. There are dedicated speech recognition systems. Other systems filter key words from a continuous flow of speech, and still others learn new words through multiple speech inputs from different speakers, thus becoming independent of the characteristic speech pattern of a single speaker. It becomes problematic for speech recognition systems when words of different meaning but with the same phonemes occur in the speech input, such as "more" and "sea".
Speech recognition systems use artificial intelligence( AI) and convolutional neural networks( CNN). Speech recognition systems are used in text entry, speech processing, driver assistance systems, voice mailboxes, computer games, dial-up devices, call centers, voice portals, IVR systems, digital voice assistants, voice dialing, and many others.