Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

"The Future of Machine Learning & JavaScript"Asim Hussain

185 views

Published on

There are many exciting things happening with AI, from which, until recently, JavaScript developers were largely shut out. But things are changing, if you can do `npm install @tensorflow/tfjs` or make an API call, you can now do AI. In this fast-paced talk, I'll open your mind to what's possible by demoing several AI-powered JavaScript apps and show you how they were built using either TensorFlow.js or easy to use AI powered APIs. You don't need a PhD in Maths, you don't need years of experience, you just need imagination and the willingness to try.

Published in: Technology
  • Be the first to comment

  • Be the first to like this

"The Future of Machine Learning & JavaScript"Asim Hussain

  1. 1. The Future of Machine Learning & JavaScript @jawacheNodeUkraine 2019
  2. 2. Asim Hussain @jawache codecraft.tv asim.dev microsoft.com
  3. 3. https://aka.ms/jawache-cda @jawache
  4. 4. @jawachehttps://www.palinternship.com/
  5. 5. @jawache
  6. 6. Asim Web Development Machine Learning This is @EleanorHaproff's slide
  7. 7. @jawache
  8. 8. TheMojifier™ @jawache
  9. 9. @jawachethemojifier.com
  10. 10. How to Calculate Emotion? @jawache
  11. 11. (1) Detect Facial Features @jawache
  12. 12. https://towardsdatascience.com/facial-keypoints-detection-deep- learning-737547f73515
  13. 13. (2) Use a Neural Network @jawache
  14. 14. Neural Networks Axon Dendrites Axons Body @jawache
  15. 15. 1 23 8.6 -0.5 2.1 Activation Function @jawache Neural Networks
  16. 16. 1 23 8.6 -0.5 2.1 x x activation(...) = -11.5 = 7.01 ->->} @jawache Neural Networks
  17. 17. Output 0 0 1 Input @jawache Neural Networks
  18. 18. 1.1 4.2 0.3 4 12 93 3 @jawache Neural Networks
  19. 19. 1.1 4.2 0.3 4 12 93 @jawache 8 - 8 = -53 Neural Networks
  20. 20. 1.1 4.2 0.3 4 12 93 @jawache - 8 = -53 8 Neural Networks
  21. 21. 0.1 9.2 0.2 4 12 93 @jawache 8 8 Neural Networks
  22. 22. @jawachehttps://azure.microsoft.com/services/cognitive-services/face/
  23. 23. https://<region>.api.cognitive.microsoft.com/face/v1.0/detect { "url": "<path-to-image>" } @jawache
  24. 24. @jawache
  25. 25. Summary @jawache
  26. 26. • Neural Networks are incredibly powerful • Conceptually, they are simple to understand @jawache Summary
  27. 27. TensorFlow, MobileNet & I'm fine @jawache
  28. 28. @jawache
  29. 29. @jawache
  30. 30. @jawache
  31. 31. TensorFlow.js @jawache
  32. 32. TensorFlow.js Train models Load pre-trained models @jawache
  33. 33. https://github.com/tensorflow/tfjs-models @jawache MobileNet
  34. 34. https://azure.microsoft.com/services/cognitive-services/computer-vision/ @jawache
  35. 35. https://codepen.io/sdras/full/jawPGa/ @jawache
  36. 36. @jawachehttps://twitter.com/ollee/status/930303340516216832
  37. 37. @jawachehttps://twitter.com/FrontendNE/status/930120267992616960
  38. 38. @jawachehttps://twitter.com/chrispiecom/status/930407801402347520
  39. 39. Summary @jawache
  40. 40. • TensorFlow.js doesn't have any dependancies • MobileNet is a simple way to analyse images • Azure Computer Vision API ❤ @jawache Summary
  41. 41. Image2Image @jawache
  42. 42. DEMO @jawachehttps://zaidalyafeai.github.io/pix2pix/cats.html
  43. 43. @jawache Generator Discriminator ✅ ❌
  44. 44. @jawache Generator Discriminator ✅ ❌
  45. 45. @jawache Generator Discriminator ✅ ✅
  46. 46. @jawache
  47. 47. @jawache
  48. 48. @jawache
  49. 49. @jawachehttps://github.com/NVIDIA/vid2vid
  50. 50. @jawachehttps://github.com/NVIDIA/vid2vid
  51. 51. @jawachehttps://github.com/NVIDIA/vid2vid
  52. 52. https://github.com/NVIDIA/vid2vid @jawache
  53. 53. @jawachehttps://github.com/hanzhanggit/StackGAN
  54. 54. Summary @jawache
  55. 55. • GANs learn to generate new images • They take a lot of compute to train • But the generator model can be run in the browser @jawache Summary
  56. 56. @jawache aka.ms/mojifier
  57. 57. Asim Hussain @jawache codecraft.tv microsoft.com

×