- Google has brought in a Hum to Search feature that will help users identify songs.
- The feature works on the latest version of Google app or by tapping on the mic of Google Search and humming.
- Google uses machine learning models which are trained to identify songs based on a variety of sources, including humans singing, whistling or humming.
Google has brought in a set of new updates for Search, Maps and Lens. One of the features Google has introduced is the Hum to Search feature. This feature will help users identify the songs stuck in their heads or as Google puts it, "solve your earworm" by just humming for 10-15 seconds.
To use the Hum to Search feature, users have to open the latest version of the Google app or tap on the mic of Google Search and say "what's this song?" or click the "Search a song" button. Alternatively, users can say "Hey Google, what's this song?" and then hum for 10-15 seconds.
Google will then show the most likely options available on the tune from which users can select the best match. The feature is currently available in English on iOS, and in more than 20 languages on Android. And we hope to expand this to more languages in the future.
"After you're finished humming, our machine learning algorithm helps identify potential song matches. And don't worry, you don't need perfect pitch to use this feature," Krishna Kumar, Senior Product Manager, Google Search wrote in a blog post.
Kumar further noted that after identifying a song users can explore information on the song and artist, view any accompanying music videos or listen to the song on your favorite music app, find the lyrics, read analysis and even check out other recordings of the song when available.
The feature works like other song identifying apps, with no music, other than what users are humming required to play in the background. Google uses machine learning models to put the Hum to Search feature in action. Google compares a song melody to a fingerprint, both of which have their unique identity.
Once users hum the melody or tune into Search, Google's machine learning models transform the audio into a number-based sequence representing the song's melody. The machine learning models are trained to identify songs based on a variety of sources, including humans singing, whistling, or humming, as well as studio recordings.