Latest inventions in natural language tools


What is speech recognition

Speech recognition (in many contextsachieve very high performance in
also known as 'automatic speechcontrolled conditions. Part of the
recognition', computer speechconfusion mainly comes from the mixed
recognition or erroneously as Voiceusage of the term speech recognition and
Recognition) is the process ofdictation.
converting a speech signal to a sequenceSpeaker-dependent dictation systems
of words, by means of an algorithmrequiring a short period of training can
implemented as a computer program.capture continuous speech with a large
Speech recognition applications thatvocabulary at normal pace with a very
have emerged over the last few yearshigh accuracy. Most commercial companies
include voice dialing (e.g., Call home),claim that recognition software can
call routing (e.g., I would like to makeachieve between 98% to 99% accuracy
a collect call), simple data entry(getting one to two words out of one
(e.g., entering a credit card number),hundred wrong) if operated under optimal
preparation of structured documentsconditions. These optimal conditions
(e.g., a radiology report), domoticusually means the test subjects have 1)
appliances control and content-basedmatching speaker characteristics with
spoken audio search (e.g. find a podcastthe training data, 2) proper speaker
where particular words were spoken).adaptation, and 3) clean environment
Voice recognition or speaker recognition(e.g. office space). (This explains why
is a related process that attempts tosome users, especially accented, might
identify the person speaking, as opposedactually find that the recognition rate
to what is being said.could be perceptually much lower than
Speech recognition technologythe expected 98% to 99%).
In terms of technology, most of theOther, limited vocabulary, systems
technical text books nowadays emphasizerequiring no training can recognize a
the use of Hidden Markov Model as thesmall number of words (for instance, the
underlying technology. The dynamicten digits) from most speakers. Such
programming approach, the neuralsystems are popular for routing incoming
network-based approach and thephone calls to their destinations in
knowledge-based learning approach havelarge organizations.
been studied intensively in the 1980sBoth acoustic modeling and language
and 1990s.modeling are important studies in modern
Performance of speech recognitionstatistical speech recognition. In this
systemsentry, we will focus on explaining the
The performance of a speech recognitionuse of hidden Markov model (HMM) because
systems is usually specified in terms ofnotably it is very widely used in many
accuracy and speed. Accuracy is measuredsystems. (Language modeling has many
with the word error rate, whereas speedother applications such as smart
is measured with the real time factor.keyboard and document classification;
Most speech recognition users would tendplease refer to the corresponding
to agree that dictation machines canentries.)



1 A B C D 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108