Hum-Based Song Search; YouTube recently revealed a new experiment for Android devices that can identify songs by humming, which is a significant improvement above Apple’s Shazam music identification tool.
According to information on YouTube’s support page, the video-sharing website is now testing a search-by-song feature for the Android app that enables users to locate a song on YouTube by humming, singing, or recording a song.
Users who are able to access the experiment can switch from the YouTube voice search to the new song search tool and hum, sing, or record a song for at least three seconds. The technology then recognizes the song and sends the user toward appropriate YouTube videos that include the song they were searching for, whether they be the official music video, user-generated content, or Shorts.
Only a limited percentage of Android users can perform searches by songs. Given that many people use YouTube to seek up music, if the tool becomes more widely available, we can see it being beneficial for many.
Also Read
Some people might be able to relate to YouTube’s most recent experiment. The feature was originally made available by YouTube’s parent firm Google in 2020 for the Google app, Google Search widget, and Google Assistant, allowing users to identify songs by humming, whistling, or singing into the microphone icon. However, it seems that the primary distinction is that the Google tool necessitates users to hum for 10 to 15 seconds in order to recognize the music.
As previously stated by Google, the feature is based on machine learning algorithms that can compare a person’s hum to the “fingerprint” or distinctive melody of a song. According to the corporation, the new YouTube test makes use of the same technology as the Google feature.
While SoundHound and MusixMatch are other music recognition programs that can identify songs by singing or humming the tune, they are less well-known than YouTube and Google. Nevertheless, we advise looking into them as well.
