decodeDynamics

As a Dynamics 365 CE Solution Architect working with IVR in the context of omnichannel customer engagement, you’ve probably come across the term ASR, which stands for Automatic Speech Recognition.
In this blog, we’ll give you a basic understanding of ASR. While it’s a deep topic, we’ll stick to the essentials.

In the world of technology, we’ve learned that computers understand a special language called binary code. Every letter or character we type or use, gets turned into a unique series of ones and zeroes.

Now, here comes an interesting question: How do computers understand our spoken words? Human speech is really complex, so how does that magic happen? Well, this is where Automatic Speech Recognition (ASR) steps in. Imagine ASR as a clever translator between our voice and the computer’s language. When we talk, ASR takes our spoken words and turns them into text (Speech to Text) and then the traditional conversion from text to binary flows as it is.

ASR is a engine which is responsible for Speech to text conversion.

A very common example of ASR engine is Siri, below image explains Siri’s ASR working.

If I ask “Hey Siri , when did Chandrayaan 3 landed on Moon”, the ASR transcripts my spoken words into Text and then internally it coverts text to binary, and then displays the results.

Another example to be called out is Alexa.

Why is ASR important: Imagine being able to talk to your devices instead of typing, especially if you’re not comfortable with reading or writing. That’s where Automatic Speech Recognition, or ASR, comes in. It’s like a smart listener for your gadgets. Instead of needing to know how to read or write, you just speak, and the technology understands you. This is really a game changer for the remote areas where literacy rate is low. Automatic Speech Recognition (ASR) serves a powerful purpose: to make technology truly accessible to everyone.

Deep Neural Network: All of the examples which we spoke above like Siri and Alexa, is uses the ASR engine which is powered by Deep Neural Network. We will talk of deep Neural Network in upcoming sessions.

To Be Published [Deep Neural Network]