Speech recognition is used for two main purposes. First and foremost dictation that is in the context of speech recognition is translation of spoken words into text. Second controlling the computer, that is to develop such software that probably would be capable of authorizing a user to operate different applications by voice [1].some of them described below:
1. People with disabilities can benefit from speech recognition programs. Speech recognition is especially useful for people who have difficulty using their hands, in such cases speech recognition programs are much beneficial and they can be used for operating computers. Speech recognition is used in deaf telephony, such as voicemail to text.
2. Individuals with learning disabilities who …show more content…
Saying something in a different tone can for example mean the exact opposite, for instance in the case of sarcasm. Although the sound is the same, the tone is different. Speech recognizers are good at recognizing sounds, but still have a hard time distinguishing the tone. Therefore is much research being done in order to detect the tone of an utterance and the emotion behind it, being anger, happiness, sadness etc. Tones are also important for certain languages, particularly Asian ones, such as Chinese and Japanese. For example, in Japanese, a word with two syllables can have different meanings, depending on whether the emphasis is on the first or second syllable, or no …show more content…
o Speaker Independent SR System
Which is a system that can recognize the speech of any speaker. This is the most difficult type in SR systems. There has to be a large sounds database for training this system.
Telephone services are open to anyone with a phone and who is willing to dial any number. Therefore a whole variety of speakers can be expected to make use of these services. In contrast to dictation systems where the speaker is known beforehand and performance can be optimized for one speaker, the recognition must be as optimal for an as large group as possible. Because of this factors which affect the voice such as race, dialect, non-nativeness, gender, age, health etc. are all variables for which an as wide range as possible should be covered by the recognition system. This makes the problem much harder and the relative performance is therefore lower than for a system for which the speaker is known beforehand. The amount of necessary training data needed is also larger, because a lot of speaker variability should be covered by the training set.
• Noise and
1. People with disabilities can benefit from speech recognition programs. Speech recognition is especially useful for people who have difficulty using their hands, in such cases speech recognition programs are much beneficial and they can be used for operating computers. Speech recognition is used in deaf telephony, such as voicemail to text.
2. Individuals with learning disabilities who …show more content…
Saying something in a different tone can for example mean the exact opposite, for instance in the case of sarcasm. Although the sound is the same, the tone is different. Speech recognizers are good at recognizing sounds, but still have a hard time distinguishing the tone. Therefore is much research being done in order to detect the tone of an utterance and the emotion behind it, being anger, happiness, sadness etc. Tones are also important for certain languages, particularly Asian ones, such as Chinese and Japanese. For example, in Japanese, a word with two syllables can have different meanings, depending on whether the emphasis is on the first or second syllable, or no …show more content…
o Speaker Independent SR System
Which is a system that can recognize the speech of any speaker. This is the most difficult type in SR systems. There has to be a large sounds database for training this system.
Telephone services are open to anyone with a phone and who is willing to dial any number. Therefore a whole variety of speakers can be expected to make use of these services. In contrast to dictation systems where the speaker is known beforehand and performance can be optimized for one speaker, the recognition must be as optimal for an as large group as possible. Because of this factors which affect the voice such as race, dialect, non-nativeness, gender, age, health etc. are all variables for which an as wide range as possible should be covered by the recognition system. This makes the problem much harder and the relative performance is therefore lower than for a system for which the speaker is known beforehand. The amount of necessary training data needed is also larger, because a lot of speaker variability should be covered by the training set.
• Noise and