Present

Gradually voice assistants are coming close to our daily life. We are surrounded by them. Mobile phones, smart devices, smart appliances all of them feature one or the other voice assistant. Alexa, Google Assistant, Siri are some of the major players.

As we are going down the road they are becoming even more smarter and customizable. Take an example of Amazon’s Alexa. You can create a skill/command and can do whatever you want. Amazon’s Alexa development kit will help you integrate with your application. With provided tools, Alexa will understand what the user is saying, what is the intent. You need to focus on your functionalities and leave the language/speech processing part to your voice assistant.

The same thing applies to other alternatives as well. Each of them has its own set of terminologies. All this may sound like a piece of jargon if you aren’t from a technical background, just ignore it if that’s the case.

Let’s deep dive, into how it works? Whenever you say/invoke your voice assistant, Alexa for instance, it starts listening to your input. Whatever the user says in the form of speech is converted into text. This is called Speech To Text (STT). This is important because computers can understand the text but not speech. Then comes the role of natural language processing (aka NLP) which is considered to be one of the most difficult dimensions of machine learning. With the use of NLP, simple plain text is used to find the actual intent or what the user means. Here the actual context of the sentence plays a vital role. For example, if the user said “eight” or “ate” they sound similar but are different. These kinds of cases are evaluated with other words in the phrase to identify what the user is saying. Once the intent is identified, processing of this intent to get a response is given to an API. Response from the API which is in the form of text is again converted into speech. This is known as Text To Speech (TTS).

Future of Voice Assistants

How easy it would be to say “Hey, book tickets for XYZ movie for tomorrow’s morning show”. Fascinating right? We suppose it’s not too long when voice assistants will be a crucial part of our lives. One of every five households already has Alexa. Today, starting from smartphones, smart home appliances to specifically made smart devices like Amazon’s Echo range this might end up replacing all our bots.