2- AI, machine learning and deep learning
ฝัง
- เผยแพร่เมื่อ 12 มิ.ย. 2024
- In this video, I discuss the difference between AI, machine learning and deep learning from a high-level point of view. I also introduce different machine learning paradigms (i.e., supervised, unsupervised and reinforcement learning) and provide suggestions about when to use traditional machine learning techniques or deep neural networks.
Deep Learning for Audio with Python GitHub (you'll find code and slides here):
github.com/musikalkemist/Deep...
Video slides:
github.com/musikalkemist/Deep...
AlphaZero: Shedding new light on chess, shogi, and Go (DeepMind article):
deepmind.com/blog/article/alp...
Interested in hiring me as a consultant/freelancer?
valeriovelardo.com/
Join The Sound Of AI Slack community:
valeriovelardo.com/the-sound-...
Follow Valerio on Facebook:
/ thesoundofai
Valerio's Linkedin:
/ valeriovelardo
Valerio's Twitter:
/ musikalkemist - วิทยาศาสตร์และเทคโนโลยี
Valerio, as a lifelong musician and amateur programmer who learned Python about 2 years ago, you have hit the sweet spot for me. Thanks for the awesome material!
I'm really glad you like the series! AI music it's such a fascinating topic, and it's difficult to find people who have the necessary computational/music theory background to tackle it at its best.
3brown1blue is great. Yeah, you would definitely go for DL libraries for a production-ready application. But implementing algorithms from scratch is the best way to really understand what happens under the hood.
i know I'm kinda off topic but do anyone know of a good site to watch newly released series online?
@Otis Kamden flixportal :)
@Solomon Orion thank you, I went there and it seems to work :) I appreciate it!
Came here to be amazed by deep learning. Ended up amazed by his way of teaching❤
Valerio, you can doing an amazing work with all your courses. I can't believe I am leaning from such quality videos for free. I am so happy that I am learning from this.
программист-любитель, изучивший
Thanks for this material! I am late to the party but I appreciate this series because it is audio focused. I think there is a lot of advancement now with AudioShake and Studio audio stem separation tools, but not a lot of hands-on tutorials. I think maybe because it requires copyrighted materials to train, and access is difficult? In any case, being comfortable with the theory and tools used to make separation tools is invaluable.
Crisp and Clear explanations . Great work
Thanks!
This is Brilliant Mr. Valerio
Thanks a ton... Valerio for this amazing tutorial.
You're welcome!
Grazie Valerio. bel lavoro come al solito!
Grazie mille :)
Thanks Valerio. Great job indeed!
awesome!!! so happy to find this deep learning videos) thank you, Valerio Velardo
Glad you like it!
Thank you! Clear and concise. 🤟
Thank you. I hope you reach amazing heights in your career :)
Thank you for the kind words :)
thanks for the video, its very informative.
Thanks bro! Very useful course! I'll have to watch.
Thank you!
Thanks.
Great !
Valero thanks a lot for this free tutorial its been really helpful. Gabriel = "Musician and Programmer"
Very interesting! Looks like fun to potentially create filters for these [i like the format so far on this series]
Thanks!
Thanks for the video mate! this is really helping me to achieve some really special project of mine.
How can I contact you . I have few doubts about my project , Im sure you can clarify them.Let me know how can I reach you
Thanks for the video
Thanks for the video Professor :)
Thx a lot for this tutorial!!!
Hi, Valerio! I'm developing a system for classification of urban noises. Do you think that classical ML algorithms would be enough or I should go to DL? I'm using the KNN but there is not much material on the internet applied to audio, so I'm confused about how it works, and you have such a nice course about DL that it maybe easier to undertand.
09:14 -> is that slack?
Great work man! Great work!
Yep, that's Slack! Glad you like this :)
You're saving my dissertation right now mate, thank you!
Great Valerio!
Thank you :)
Can deep learning be used to detect the quality of singers audio files ? inshort to detect whether the song is gonna going to get hit or flop ??
Apart from CNNs, RNNs (and lately GANs) are there any other architectures being used with audio? Also, how different is to apply a deep learning model that is trained for a specific task (etc. audio classification), in real time audio? Would be good to know how I could use deep learning to build an audio VST plugin in JUCE for example (where you work with buffers etc.).
Good question! Most types of network architectures have been used for audio/music. To name a few beyond the ones you mentioned: simple multilayer perceptrons have been initially used for many audio classification problems, Variational Auto Encoders have been used for music generation and music recommender systems, and Transformers are quite fashionable for generative music now.
@@ValerioVelardoTheSoundofAI Interesting. I've seen that autoencoders are used for audio denoising/restoration, also lately 1d CNNs are used for sequenced data (like audio), don't know if anything applied in music though. You probably missed my last question as I edited my initial comment just by the time you were replying back to me (check above). Thanks for the feedback!
Hello, can you recomend me a book that u used for this course? :) Thanks
Which do you think is better for DL, Tensoflow or Pytorch?
I personally prefer TF/Keras because it has a great environment for production.
PyTorch is great for research purposes.
@@ValerioVelardoTheSoundofAI Thanks for your quick reply, I've just started your series and am on video 5 now. Have liked and subscribed to your channel!
To start, I know nothing of python and find it doesn't seem to click , been using MATLAB for school and just much more friendly for my brain I guess. That being said I know Python is used the longest so maybe I'll learn it but wanted to share a thought. I ended up at some videos further down, my favorite so far was the one about MFCC and speech signals. I don't know how relevant this will all be to me in the long run but I wanted to say I find your teaching and videos to be some of the easiest and most fun to follow so even if I don't follow the code I'm still looking forward.
Thank you JD - If you'd like to work in indusrtry in DL/ML, Python is a must. Try to give it another try ;)
@@ValerioVelardoTheSoundofAI by the time I'd be in industry won't be much left to develop... lol I'm going to try and follow your Python but if I remember correctly you aren't using Windows so everything is different.
@@JD-jj5ue Python is cross-platform. What I do on Linux is also valid on Win.
Is the intro sound from PS3 starup 😅?
I'm not aware of the PS3 startup sound. This is just the sound of an orchestra tuning.
3:40 Optimize actions to achieve goals? XD So some times humans are irrational that's what you're meaning?!?!? XD hahahahah xD