March 17, 2018

Voice Recognition – AI for Audio

Voice recognition is a prominent technology that is being used in more and more places today.

There is voice recognition for recognizing the speaker. This type of voice recognition is used for security (physical access and access to files in computer network). This is also evolving to give specific instructions/information to the speaker as well.

There is voice recognition for translation, as well. This allows for written transcripts of audio events. The benefits are not limited to ADA compliance but useful for translating into other languages and to have a searchable transcript of an event.

In both cases, artificial intelligence (AI) is utilized to have the computer learn and improve on its success of voice recognition.

While everyone has heard of Siri or Alexa for voice recognition, there are innovative uses far beyond the home or smartphone. One recent example is in forensics. A voiceprint can be created in less than two minutes of speech. Crime experts are now using voiceprints for identity of voices from wire taps. This application is unique in that it uses the sound of one’s voice rather than the shape or pattern. That means that a voice match can be found regardless of the language or accent of the person.

At UAT, we implement both of these types of voice recognition in our classes, such as AI, machine learning, and deep learning. Some of the examples of what our students have done are:

  • Coding by Voice – This allows visually impaired or those driving to speak their code. The cool part is that when you write a line a code then you can say “check” and it will make sure your syntax is correct or make suggestions on corrections.
  • Speech Analysis for Intent – When you send an email, write a blog, or read a news article, the content can be analyzed for intent – positive or negative. This could be used to find only positive new stories. This way, you could filter out the negative stories or blog posts. It could also be used in emailing your boss about a problem. This is important because in the written word, some meaning is lost without the actual interaction and this would help to insure that the message you want to convey has the proper intent.

Voice recognition is just one of the technologies that UAT students master – because we excel at implementing new technology as it occurs. UAT is truly a University of Advancing Technology.


Leave A Comment