Google is expanding on its Search platform with a new experimental pronunciation feature as well as a visual definition function.
The search giant announced the new features in a blog post, noting that people often ask language-related questions. Now Search will go above and beyond simple spelling and audio that demonstrates how to pronounce a word.
First up, Search will now be able to teach users how to pronounce a word. Previously, when you looked up a word like ‘quokka,’ you’d have an option to play audio of the pronunciation. However, users can now respond by saying the word into their phone’s microphone. Google will provide feedback on your pronunciation and help you correct any issues.
To start, the feature is rolling out now in American English. Google says it will come to Spanish soon and more languages after.
Search performs the pronunciation feature by using sound recognition technology to separate spoken words into individual soundbites. Then, Search uses machine learning to cross-reference those sounds bites with how the word should be pronounced. If it doesn’t line up, it issues feedback based on that soundbite.
The other major feature rolling out now is visuals in definitions. Google wants to add additional context to a word, which can be useful with words that have multiple meanings, sound similar or aren’t common.
To add that context, Google will display images along with definitions. However, since some words are hard to explain with a picture, the company will start with nouns and expand from there. Google says the feature will be available in English first and available across all language translations.
You can learn more about the new features in this Google blog post.