DSP Background - Deep Learning for Audio Classification p.1

thank you)

Hey what does it mean dealing with mel spectrograms (128,216,3). By using 3 windows length 93ms,46ms and 23ms and in the end they have write 128,216,3 what does 3 shows here??

Halfway into this video and already love your no BS approach. You're my new "that guy". Solid

explained really clearly!!! thanks a lot.

what an angel. What a great playlist!! Thank u

Hey bro you look like somewhat neville longbottom

Dude awesome!

This was amazing. Thank you!

Still in the video but this is one of the best explanations of STFT i've seen (I'm coming from amateur radio background and have been using these for years without full understanding how these contribute to the spectrogram views we see)

What's that thumping noise in the video?

You need a highpass filter for your mic though, the banging sub noise is all over.

Hello I hope you are well, I followed your videos on deep learning for audio classification and it was very interesting thanks for everything.
but please ask her something:
if i want to create a machine learning model for transcribing (audio -> written) in a brand new language, like an african language for example, how should i proceed.
thank you

very instructive talks.

Top video. Thank you!

Great videos Seth. Thanks a lot!

How to deal with and overcome overlapping sounds?

what is your tensorflow version

ehh.... you should remove your camera video it is distracting . It fine at the beginig but throwing around the screen it really terrible idea.

Hi, This is a cool video and has really helped me understand ML in audio, but could you recommend a good source for audio source separation or other audio ML collabs?
Thanks!

I’m an Audio Engineer and this is a VERY COOL Channel! Keep creating content! 🤓

😍😍

That deep noise just threw me off. Maybe I'm the only one but it was very distracting.

hello , I have problem in below line :
rand_index=np.random.randint(0,wav.shape[0]-config.step)
type of error :
ValueError: Range cannot be empty (low >= high) unless no samples are taken

can you help me?

Hi! I as a fan of guitar amp profiling do you think it could be a game changer for next Kemper or guitar amp plugins in real time?
Thanks :)

Can you suggest some good course to learning deep learning using tensorflow for beginners?

This is fantastic! Thanks for sharing

Would I need a different example from this in order to detect a distinct human sound such as a sneeze or cough?

Can we use this example for human speech recognition?

amazing work..keep it up...

Thank you, this is a very good job! 👨🏻‍🏫

Nice. Any other resources you could suggest for audio signal processing and deep learning?

Hey, Can you please direct me to the haythemfayek's blog because the link that you have provided seems to have expired. Thanks in advance.

In video you mentioned nyquist frequency = highest frequency...A simple google search gave me : nyquist frequency = 2*(highest frequency)...still very good video.

Very helpful, it is like ELI5

Not to split hairs, but when you are talking about the FFT you are showing the equation for the continuous-time Fourier transform.

Excellent video! Great explanations of a very challenging topic.

Very useful video, thanks!

Was verrrrrrrrrrry helpful. You did a good job in explaining. Thank you very much

Thanks for the lecture! Just a quick question, can it be used to classification on human voices?

Thanks, Man. I'm working on Audio Denoising using Deep Learning can you make a video or something relevant to this ??

Note that the FFT window size DOES NOT allways have to be power of 2, e.x numpy does not use radix-2 method, so it might be useful if you want accurate analysis on specific frequencies

I have an application that I need to monitor a (fairly noisy) machine for proper operation, it changes sound when it fails, Would Deep Learning for audio would be a good fit?

hi bro, it seems ur voice is not clear.... i cant get most of the important things

Hello, please explain me the final step how 13 coefficients are selected out of 26 coefficients,all 26 filter bank energies are prominent features.