What’s ASR? It stands for Automatic Speech Recognition, where computers “listen” to what you say and then respond as if you’re having a conversation. Describing it like that may make it sound like something out of science fiction, but ASR is actually something that you’ve almost certainly used before – probably fairly recently.

Where Can ASR Be Found?

Automatic speech recognition programs are used in lots of ways. The technology has been around since the 1950s, and some disabled individuals have been using it since the 1980s, but most people today are familiar with it from one of two places:

Answering Systems - Whenever you call your cable company, bank, or local Walgreens, you’re greeted by a computerized answering system that uses voice prompts to get you where you want to go. This is a type of ASR called Direct Dialog.

iPhone's Siri and Windows Phone's Cortana - We’ve all seen those commercials of famous people asking Siri or Cortana various questions, right? Well, they’re using a type of technology called Conversational ASR, which not only comes preprogrammed with a large list of common questions, but also has the ability to “learn” your speech patterns and common requests over time.

How the ASR Technology Works

ASR is a complicated technology that continues to improve and evolve all the time, but the basics of how it works haven’t changed for years -

• The user speaks into a microphone.

• Those sounds are translated into a wave form.

• The wave form is then broken down into phonemes – the sounds that make up language.

• Through looking at the first phoneme and calculating the likelihood of which phoneme will come next, the program is able to recognize entire words and phrases in near real-time.

• Conversational ASRs also do these calculations with entire words as they “learn” user tendencies, so if you frequently ask “What restaurants are close by?” the program will be ready to look for restaurants and offer that as an option as soon as you say “what.”

Automatic Speech Recognition Isn’t Perfect – But It’s Getting There

ASR proponents will tell you that the technology is accurate 96 percent of the time “under ideal circumstances.” It’s that caveat that does most programs in, though, because humans are rarely using their ASR devices in ideal circumstances. If music is playing nearby, cars are whizzing past, or other people are talking, those things lower the accuracy significantly. And even the most advanced ASR program out there can’t make up for a shoddy microphone.

Still, it’s a technology that continues to improve and become more and more ubiquitous, so you need to know how to use it. If you want to learn more about automatic speech recognition, check out the infographic from West Interactive.

How Apple's Siri and Microsoft's Cortana Works Technically on Your Smartphone (Infographic)

How Apple's Siri and Microsoft's Cortana Works Technically on Your Smartphone (Infographic) -
Wanna Embed This Infographic in Your Blog Post? Copy and Paste the Code Given

Post a Comment Disqus