Ask HN: Machine learning resources for audio processing • Hacker News Board

Re: Ask HN: Machine learning resources for audio processing

#11

Post by sdenton4 » Thu, Apr 18, 2019, 5:10 PM UTC

The folks behind audio set have been working on general audio event detection for some years now, I believe.

https://research.google.com/audioset/

There's a huge amount to discuss in the audio domain... But for a starting place, using ResNet on spectrograms to build a binary classifier is a good place to start.

Re: Ask HN: Machine learning resources for audio processing

#12

Post by Tangokat » Thu, Apr 18, 2019, 5:13 PM UTC

I don't know if this is off topic but would it be possible to remove the sound of mechanical keyboards with ML in realtime from a VOIP stream? Sell the technology to Discord and profit.

Re: Ask HN: Machine learning resources for audio processing

#13

Post by williamsmj » Thu, Apr 18, 2019, 5:17 PM UTC

I think the slides/recording of this excellent Spotify talk will be posted shortly: https://qcon.ai/qconai2019/presentation/deep-learning-audio-....

Re: Ask HN: Machine learning resources for audio processing

#14

Post by zneveu » Thu, Apr 18, 2019, 5:43 PM UTC

am807 wrote: post #4

Just found this thread on the fast.ai forum yesterday that may help: https://forums.fast.ai/t/deep-learning-with-audio-thread/381...

I just found this too! Great thread for the new Kaggle Freesound competition.

Re: Ask HN: Machine learning resources for audio processing

#15

Post by telesilla » Thu, Apr 18, 2019, 5:59 PM UTC

telesilla wrote: post #8

aubio and librosa are two excellent MIR (music information retrieval) tools I can recommend from personal use. They can both be implemented for real-time audio using pyaudio or similar. https://aubio.org/doc/latest/ https://librosa.github.io/librosa/

To append to my own comment, something I haven't tried myself yet but I'm planning to: Urban Sound Classification with Neural Networks in Tensorflow

https://www.kdnuggets.com/2016/09/urban-sound-classification...

Re: Ask HN: Machine learning resources for audio processing

#16

Post by colin_fraizer » Thu, Apr 18, 2019, 6:21 PM UTC

Is anyone working on removing underscoring—by which I mean background music, not underlining—from video with dialogue?

Re: Ask HN: Machine learning resources for audio processing

#17

Post by jamesb93 » Thu, Apr 18, 2019, 7:03 PM UTC

This depends if you're interested in creative applications or analytical (MIR) ones. The two fields share a lot of techniques, but the way they are used is wildly different.

Re: Ask HN: Machine learning resources for audio processing

#18

Post by iagooar » Thu, Apr 18, 2019, 8:08 PM UTC

Contact the founder / maker of Auphonic.com - he's a super nice and clever guy who does this kind of stuff for a living. He'll definitely point you into the right direction.

Re: Ask HN: Machine learning resources for audio processing

#19

Post by alphagrep12345 » Thu, Apr 18, 2019, 8:35 PM UTC

Tangokat wrote: post #12

I don't know if this is off topic but would it be possible to remove the sound of mechanical keyboards with ML in realtime from a VOIP stream? Sell the technology to Discord and profit.

Is this a big problem? I thought people loved mechanical keyboard sounds.

Re: Ask HN: Machine learning resources for audio processing

#20

Post by wpietri » Thu, Apr 18, 2019, 8:38 PM UTC

alphagrep12345 wrote: post #19

Tangokat wrote: post #12

I don't know if this is off topic but would it be possible to remove the sound of mechanical keyboards with ML in realtime from a VOIP stream? Sell the technology to Discord and profit.

Is this a big problem? I thought people loved mechanical keyboard sounds.

People love their own, I'm sure. I've never met anybody who puts on a tape of it, though.