Published
Wed, Apr 24, 2024, 12:10
- Launched earlier this month, the website currently contains 1,152,633 records in its database.
A new online digging tool recommends users electronic music tracks based on similarity.
Launched earlier this month by the brains behind popular YouTube channel hurfyd, cosine.club calls itself a "music similarity search engine." Users simply input a track title into the search bar and, if recognised, the engine returns multiple results. The database currently contains 1,152,633 tracks.
For example, if you input Joey Beltram's "Energy Flash," you'll receive DJ's Guess Who's "Well Done," Gudrun Gut's "Firething (Radio Mix)" and Clementine's "Groove Me," among others. All the tracks are ranked by compatibility and can be streamed on the platform via YouTube.
According to the About section, cosine.club "repurposes a machine-learning model originally designed to classify music genres." It works by calculating "a cosine similarity metric between the source and all other tracks in the dataset to find the closest matches."
Last year, UK label Timehri Records developed a similar machine-learning tool, Timehri Dig Assistant, using Discogs URLs.
Browse cosine.club.
