Been keeping an eye on the team for a while, as they're quite visible in the Berlin music tech ecosystem. Very cool to see them launch an app that uses computer vision to recommend music. No easy task, and will be interesting to see how the app gets developed further in the future! Question for the makers: how do you build up the data to train your neural networks?

How it determines our mood ? Based on past searches !

Hello @basgras, many thanks for hunting Cyanite PLAY! Hey Product Hunters 🖖🏻 2 months ago we have launched CYANITE (https://www.producthunt.com/post...), an AI tool that visualizes emotions in music. Today we are very happy to continue with Cyanite PLAY, the corresponding iOS app that gives music recommendations for your videos. The reason for Cyanite PLAY is the problem we've been dealing with for quite some time, that you often don't know what to look for in music or how to translate what you want to hear in a certain situation into keywords. Especially in times when you have millions of songs in your pocket. That's why we asked ourselves: How can we use new ways of discovering music, away from big playlists and algorithms based on our own listening habits? And how can we use new technologies like Computer Vision, AI and Machine Learning? Cyanite PLAY is our first answer to these questions. It's basically a way to discover new music without having to search for it. You let your situation, your video and your mood speak. To your question Bas on how we train our neural networks: The dataset comes from our own app Groovecat, which we released a year ago. It's an app to capture your Music Moments, the moment a song hits the sweet spot. With the Music Moments, our system has learned which songs work together with an emotion in a certain situation. We have derived our insights from these Music Moments and trained our models with them. Currently, our computer Vision covers 10 different settings, more will follow soon. Because we put so much commitment and love into the development of Cyanite PLAY, we are looking forward to your feedback from different angles and backgrounds. If you have suggestions, improvements and ideas: We are all ears! 👏🙏🏻🎧 #tl;dr ▶︎ iOS app that recommends music based on your video and mood. ▶︎ For music listeners and creators of visual content alike. ▶︎ Technology is based on Computer Vision and a smart music database.