7 Like 0 Dislike
3 hours of Computery's music: https://www.youtube.com/watch?v=vvusi8EWzPc The training data was all composed by Doug McKenzie at http://www.bushgrafts.com/jazz/midi.htm. Here's some of those pieces of music converted to images: https://www.youtube.com/watch?v=6iPZE20WACk Part 1 ("3 neural nets BTPTBJM"): https://www.youtube.com/watch?v=uiJAy1jDIQ0 Want to chat about science, technology, engineering or math? Then go here: https://discord.gg/dxCBaYn What about non-STEM stuff, like TWOW or BFDI or anything else carykh-related, then you can go here: https://discord.gg/ZVQMa8U My patreon: https://www.patreon.com/carykh Raw MIDI, .wav, .mp3, and .txt files: https://drive.google.com/drive/folders/0B98DkR2AsGZaYk9JNFJScWRnMTQ?usp=sharing Original human-composed jazz MIDI files: http://www.bushgrafts.com/jazz/midi.htm Andrej Karpathy's LSTM: http://karpathy.github.io/2015/05/21/rnn-effectiveness/ https://github.com/karpathy/char-rnn HyperGAN: https://github.com/255BITS/HyperGAN PixelCNN: https://github.com/openai/pixel-cnn Hazel Cricket: https://www.youtube.com/channel/UCasKvuIk59T7SUJr60auxqg AI-generated baroque music video: https://www.youtube.com/watch?v=SacogDL_4JU Teaching my computer to give me friends (video about convolutions): https://www.youtube.com/watch?v=p_7GWRup-nQ My twitter: https://twitter.com/realCarykh Background music (human composed): Great Days by Joakim Karud http://soundcloud.com/joakimkarud Music provided by Audio Library https://youtu.be/5lhZRunuJTs Almost Original (Instrumental) by Joakim Karud http://soundcloud.com/joakimkarud Music provided by Audio Library https://youtu.be/r20_9c0fzGk
I put the word "evolve" in there because you guys like "evolution" videos, but this computer is actually learning with gradient descent! All music in this video is either by Bach, Mozart, or Computery. GizmoDude8128 wins a prize for figuring out that 100101 in base 2 is 37 in base 10 the fastest! (Question inspired by fixylol) Andrej Karpathy's blog post on RNNs: http://karpathy.github.io/2015/05/21/rnn-effectiveness/
http://www.ted.com Stephen Wolfram, creator of Mathematica, talks about his quest to make all knowledge computational -- able to be searched, processed and manipulated. His new search engine, Wolfram Alpha, has no lesser goal than to model and explain the physics underlying the universe. TEDTalks is a daily video podcast of the best talks and performances from the TED Conference, where the world's leading thinkers and doers give the talk of their lives in 18 minutes. Featured speakers have included Al Gore on climate change, Philippe Starck on design, Jill Bolte Taylor on observing her own stroke, Nicholas Negroponte on One Laptop per Child, Jane Goodall on chimpanzees, Bill Gates on malaria and mosquitoes, Pattie Maes on the "Sixth Sense" wearable tech, and "Lost" producer JJ Abrams on the allure of mystery. TED stands for Technology, Entertainment, Design, and TEDTalks cover these topics as well as science, business, development and the arts. Closed captions and translated subtitles in a variety of languages are now available on TED.com, at http://www.ted.com/translate. Watch a highlight reel of the Top 10 TEDTalks at http://www.ted.com/index.php/talks/top10
Stephen Wolfram introduces the Wolfram Language in this video that shows how the symbolic programming language enables powerful functional programming, querying of large databases, flexible interactivity, easy deployment, and much, much more. To learn more about the Wolfram Language, visit http://www.wolfram.com/language/ For the latest information visit: http://reference.wolfram.com/language http://www.wolfram.com
This is the result of a project I worked on for CS224D with Aran Nayebi. The idea is to design a neural network that can generate music using your music library as training data. While I believe we could probably improve upon this model significantly, it serves as a good proof of concept for showing that it is indeed possible. We use Long Short-Term Memory (LSTM) networks to model hidden recurrences over time. For this demo, we used a network that was trained on a variety of Madeon songs. UPDATE: Due to popular demand, we've released the source code on GitHub. Check it out here: https://github.com/MattVitelli/GRUV Happy training! :D
Dabbling with GameMaker led me to look for procedurally generated music software. After trying a few things, Wolfram Tones won out, mainly for the variety within its output.
Examples of music begin at 1:42.
A worldwide, royalty-free licence to use WT within commercial indie-games can be requested using the 'contact us' link within their website.
Backdrop to the video is a pre-pre-pre-pre-alpha-pre-pre of a small growth of code I've prodded until it did a thing.