Subscribe
Sign in
Musiio Tag — Tag songs quickly using Artificial Intelligence
Musiio Tag is an AI-powered audio analysis tool that tag songs without using metadata. Upload a YouTube link or mp3/wav file, and the AI will ‘listen’ to the audio, analyze through its neural networks and assign tags such as genre, mood and key within seconds.
14
Replies
David Reinberger
Very cool! I wish streaming apps would use such an advanced ML to suggests tracks, I'm very picky :D Also, for DJs use, if it could distinguish between electronic music genres and their sub genres - pure awesomeness.
Share
@david_reinberger Hi David, you’re absolutely spot on about the use cases for both. At the end of the day, we want to harness the power of AI to assist DJs to create their setlists effortlessly and streaming services to help their listeners discover great music.
Amazing product!
Would love to know what model you're using for the entity extraction from songs.
Hey Product Hunters, Fah from Musiio here! 👋🏼
Turn yourself into a music expert and name the key of a song in seconds. 🤓 Or go pro-DJ and find out the BPM of all your setlist tracks super quick… 😎 We’re actually a B2B search company for Audio but these are just a few of the experiments people have used our demo for.
For all audio search, we make large catalogues of music searchable and manageable.
One of the ways we do this is using automated tagging. Gone are the days of arduously assigning genre, BPM, key, mood indicator and more to individual tracks. Musiio can assign the 8 tags you see in our demo to over 1m songs a day, with 90%+ accuracy.
Now, this unscalable task is fully automated and you can search for a happy/rock/e minor song in your catalogue in seconds. Do we stop here? Absolutely not, we also have AI-powered search and AI-powered playlisting - both of which run independently to the tagging system.
But the tagging system is a nice way to see what we are about, and try it for yourselves.
Users of our Product include:
- Streaming services
- Production music/sync companies
- Record labels
- Online music stores
Try the AI for yourself. We’d love to hear your feedback! 😊
Hey @glennf__ff, no we don’t keep any user data nor monetize them. We’re a B2B company that focuses on delivering customized solutions for our clients with their music catalogues. We provide them with the knowledge capability to unlock the full potential in their database (be it tagging, search or playlisting) in a scalable manner and that’s our value add.
Tried 'Overdose' by Wolfgang Gartner and it identified it as Electronic and Pop - Amazing! Love that it can also identify Key, BPM, energy, audio type, mood...
Don't know any other tech products that do this. You guys are on to something here.
Oooooh loving this! Tried it out with a pop song (Better Now by Disco Hue) and it came back pretty accurate!
Audio Type: Music
Genre: Pop (91%)
Vocal: High Presence
Quality: High
Key: E♭ minor, F♯ major
Tempo: 101 bpm, Small Variance
Energy: Moderate, Medium Variance
Emotional Spectrum: positive
Mood: Happy (70%), Romantic (48%)
Love that they picked up 2 keys!
@amelia_amari Thanks for trying it out Amelia! Great to hear that the AI did pretty well on that track!
Wow! Brilliant technology! Tried the demo with Sam Smith and Normani's Dancing With A Stranger and it was extremely on point.
Audio Type: Music
Genre: Pop (100%)
Vocal: High Presence
Quality: High
Key: F minor, A♭ major
Tempo: 103 bpm, Small Variance
Energy: Moderate, Small Variance
Emotional Spectrum: positive
Mood: Romantic (82%), Relaxed (48%)
Completely agree with the keys....especially coming from someone with perfect pitch :p
This is one of the greatest products I've ever seen and I can't wait to see how the product grows and evolves from here on. Loved that the AI even managed to pick out keys that shared a similarity in harmonic make-up. Solid stuff.