$30 off During Our Annual Pro Sale. View Details »

Dreams, Drugs and ConvNets

Dreams, Drugs and ConvNets

Artefacts of human and artificial cognition.
at Studenckie Koło Naukowe Neurobiologii, the University of Warsaw

Piotr Migdał

March 02, 2017
Tweet

More Decks by Piotr Migdał

Other Decks in Science

Transcript

  1. Dreams, Drugs & ConvNets Piotr Migdał, PhD deepsense.io / freelancer

    http://p.migdal.pl/ 2 Mar 2017, Studenckie Koło Naukowe Neurobiologii UW
  2. None
  3. None
  4. Deep Learning progress • image recognition, neural style, word analogies,

    per-char translations, playing ATARI games, Go, [no idea what’s next] • fast-paced (more than my quantum physics PhD):
 6 month ago a breakthrough, now a baseline • (no questions about Singularity please!)
  5. Artificial neural networks • Multidimensional arrays • Simple matrix operations

    (add, multiply) • Floats (not spikes) • Simple activation functions (sigmoid, ReLU) • Cost function and back-propagation • A lot of weights (e.g. 100M)
  6. Many layers https://research.facebook.com/publications/deepface-closing-the-gap-to-human-level-performance-in-face-verification/

  7. Convolutions http://setosa.io/ev/image-kernels/

  8. Patterns activating channels https://blog.keras.io/how-convolutional-neural-networks-see-the-world.html

  9. Big questions • creating general AI or sentient beings •

    transferring our minds (mind uploading) “I mean, if we're able to save even just a small piece of ourselves, why wouldn't we do that?” - SOMA https://www.youtube.com/watch?v=BZTfi1jv-EE
  10. Flesh vs machines:
 practical questions • Cross-inspiration: • Machine Learning

    to Human Cognition • Human Cognition to Machine Learning • Common abstractions • Learning in general • Hierarchical signal processing
  11. NSFW, 18+, etc

  12. http://www.techrepublic.com/article/why-microsofts-tay-ai-bot-went-wrong/ Tay.AI (+ 4chan)

  13. The biggest lie of every
 parent, pet owner
 and neural

    network trainer: “I’ve never shown it that, it must have learned it by itself!”
  14. Stereotypes http://p.migdal.pl/2017/01/06/king-man-woman-queen-why.html

  15. If you’ve never had a trip… https://youtu.be/cPKq7JuQDvg?t=256

  16. If you’ve never had a trip… https://youtu.be/z7_U6y8kJaY?t=72

  17. Deep dreams: forcing ConvNet
 to see things https://research.googleblog.com/2015/06/inceptionism-going-deeper-into-neural.html

  18. • What should I use:
 THC, LSD, DMT? • A

    good GPU! http://alexgrey.com/shop/st-al.html
  19. Grocery trip https://www.youtube.com/watch?v=DgPaCWJL7XI

  20. Forcing networks to see things https://research.googleblog.com/2015/06/inceptionism-going-deeper-into-neural.html

  21. Neural style transfer also video: https://www.youtube.com/watch?v=Khuj4ASldmU

  22. With own photos to play with https://deepart.io/ or https://prisma-ai.com/

  23. What is sexual? http://blog.clarifai.com/what-convolutional-neural-networks-see-at-when-they-see-nudity/#.WLhGmRIrLLh

  24. What is sensual? http://blog.clarifai.com/what-convolutional-neural-networks-see-at-when-they-see-nudity/#.WLhGmRIrLLh

  25. And this? (yes, ConvNet is also fooled)

  26. Sketch to picture http://affinelayer.com/pixsrv/

  27. Sketch to picture: let’s get rogue! http://affinelayer.com/pixsrv/

  28. https://raw.githubusercontent.com/Newmu/dcgan_code/master/images/faces_arithmetic_collage.png https://github.com/255BITS/HyperGAN Picture analogies

  29. https://open_nsfw.gitlab.io/ Image generation

  30. Learning what is important • Object consistency • Toddler needs

    to learn, so do ConvNets http://blogs.scientificamerican.com/illusion-chasers/what-little-babies-see-that-you-no-longer-can/
  31. Uncertainty and learning http://www.nature.com/articles/ncomms11609

  32. Retrain your own network:
 McCollough effect http://www.michaelbach.de/ot/col-McCollough/

  33. Trypophobia
 (don’t google it!) visit here instead: https://en.wikipedia.org/wiki/Trypophobia

  34. Trypophobia detector https://github.com/grzegorz225/trypophobia-detector (a 5-day workshop was enough for 88%

    accuracy!)
  35. And if you
 don’t want to code… http://www.sciencemag.org/news/2015/11/pigeons-spot-cancer-well-human-experts https://www.youtube.com/watch?v=flzGjnJLyS0

  36. Questions? • If you want to learn: • http://lumiverse.io/series/neural-networks-demystified •

    https://gist.github.com/stared/7de2908b9bcba01c39ee3c591875a23c • http://www.deeplearningbook.org/ • Keras (on top of TensorFlow) in Jupyter Notebook • https://medium.com/@mateuszsieniawski/keras-with-gpu-on-amazon-ec2-a-step-by- step-instruction-4f90364e49ac#.4i1norivh • https://www.erowid.org/ http://p.migdal.pl/ + I do like emails: pmigdal@gmail.com