Recommended
More Related Content
What's hot
What's hot (20)
Similar to Intro To Convolutional Neural Networks
Similar to Intro To Convolutional Neural Networks (20)
Recently uploaded
Recently uploaded (20)
Intro To Convolutional Neural Networks
- 1. Intro To Convolu,onal Neural Networks Mark Scully datapraxis.com
- 2. Why CNNs?
- 7. Facial Recogni,on Y. Taigman, M. Yang, M. Ranzato, L. Wolf, DeepFace: Closing the Gap to Human-‐Level Performance in Face Verifica,on, CVPR 2014
- 9. Colorize Black & White Images h@p://richzhang.github.io/coloriza,on/
- 10. Style Transfer h@p://genekogan.com/works/style-‐transfer/ Mona Lisa restyled by Picasso, van Gough, and Monet
- 11. Generate An Image From A Sketch h@ps://affinelayer.com/pixsrv/
- 12. ImageNet Challenge Alexnet Li Fei-‐Fei: ImageNet Large Scale Visual Recogni,on Challenge, 2014
- 13. ImageNet Challenge ILSVRC+ ImageNet Classification error throughout years and groups Li Fei-‐Fei: ImageNet Large Scale Visual Recogni,on Challenge, 2014
- 14. Alexnet Architecture -‐ 2012 Input Conv Relu Pool Conv Relu Pool Conv Relu Conv Relu Conv Relu Pool FC Dropout FC Dropout ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 FC 1000
- 15. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012
- 16. ImageNet Challenge Alexnet Li Fei-‐Fei: ImageNet Large Scale Visual Recogni,on Challenge, 2014
- 17. Tradi,onal Approach To Image Classifica,on Input Image Hand Extracted Features Classifier Object Label
- 18. Issues • Who makes the features? – Need an expert for each problem domain • Which features? – Are they the same for every problem type? • How robust are these features to real images? – Transla,on, Rota,on, contrast changes, etc.
- 19. Are these pictures of the same thing?
- 20. Features Are Hierarchical • A squirrel is a combina,on of fur, arms, legs, & a tail in specific propor,ons. • A tail is made of texture, color, and spa,al rela,onships • A texture is made of oriented edges, gradients, and colors
- 21. Image Features • A feature is something in the image or derived from it that’s relevant to the task • Edges • Lines at different angles, curves, etc. • Colors, or pa@erns of colors • SIFT, SURF, HOG, GIST, ORB, etc
- 22. Edges
- 23. Ideally We’d Learn Features Input Image Output Label
- 24. Ideally We’d Learn Features Input Image Output Label CNNs
- 25. What is a Neural Network? • Perceptron is biologically inspired • A mental model for interpre,ng the math h@p://cs231n.stanford.edu/index.html
- 26. Perceptron 1 x1 x2 x3 xm Σ Output Ac,va,on Func,on Sum w0 w1 w2 w3 wm Weights Inputs
- 27. Perceptron 1 x1 x2 x3 xm Σ Output Ac,va,on Func,on Sum w0 w1 w2 w3 wm Weights Inputs wi xi i=0 m ∑ = w0 x0 + w1x1 + w2 x2 +...+ wm xm
- 29. Training: Upda,ng Weights 1 x1 x2 x3 x4 Σ Output Ac,va,on Func,on Sum w0 w1 w2 w3 w4 Weights Inputs Error = Output -‐ Target
- 31. Deep (Mul,-‐Layer) Neural Network
- 32. Backpropaga,on • Error propagates backward and it all works via (normally stochas,c) gradient descent. • (wave hands)
- 34. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012
- 35. CNN Layer Architecture Input Convolu,on Nonlinearity Pooling (op,onal) Dropout (op,onal)
- 36. CNN Layer Architecture Input Convolu,on Nonlinearity Pooling (op,onal) Dropout (op,onal)
- 37. Input: Pixels Are Just Numbers h@ps://medium.com/@ageitgey/machine-‐learning-‐is-‐fun-‐part-‐3-‐deep-‐learning-‐and-‐convolu,onal-‐neural-‐networks-‐ f40359318721
- 38. CNN Layer Architecture Input Convolu,on Nonlinearity Pooling (op,onal) Dropout (op,onal)
- 39. Goals • Need to detect the same feature anywhere in an image • Reuse the same weights over and over • What we really want is one neuron that detects a feature that we slide over the image
- 40. Neuron = Filter • Act as detectors for some specific image feature • Take images as inputs and produce image like feature maps as outputs
- 41. Convolu,on • Like sliding a matrix over the input and performing dot products • It’s all just matrix mul,plica,on
- 42. Convolu,on
- 43. Convolu,on
- 44. Filters (or Kernels) Sharpen
- 45. Filters (or Kernels) Box Blur
- 46. Filters (or Kernels) Edge Detec,on Feature Map
- 48. CNN Layer Architecture Input Convolu,on Nonlinearity Pooling (op,onal) Dropout (op,onal)
- 49. Nonlinearity
- 50. CNN Layer Architecture Input Convolu,on Nonlinearity Pooling (op,onal) Dropout (op,onal)
- 52. Alexnet Architecture 3x3 stride 2 Max Pooling
- 53. Pooling • Allows us to look at more of the image • Max, sum, and L2 pooling • A type of downsampling
- 54. CNN Layer Architecture Input Convolu,on Nonlinearity Pooling (op,onal) Dropout (op,onal)
- 55. Alexnet Architecture -‐ 2012 Input Conv Relu Pool Conv Relu Pool Conv Relu Conv Relu Conv Relu Pool FC Dropout FC Dropout FC 1000 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 Dropout
- 56. Dropout h@p://cs231n.github.io/neural-‐networks-‐2/ • Randomly disable some neurons on the forward pass • Prevents overfiong
- 57. Let’s Predict Something! • We have all these features, how do we learn to label something based on them?
- 58. Alexnet Architecture -‐ 2012 Input Conv Relu Pool Conv Relu Pool Conv Relu Conv Relu Conv Relu Pool FC Dropout FC Dropout FC 1000 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 Fully Connected
- 59. Fully Connected Layers • Each neuron is connected to all inputs • Standard mul,layer neural net • Learns non-‐linear combina,ons of the feature maps to make predic,ons
- 61. Alexnet Architecture -‐ 2012 Input Conv Relu Pool Conv Relu Pool Conv Relu Conv Relu Conv Relu Pool FC Dropout FC Dropout ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 FC 1000
- 62. Which Class Is It Again? • FC-‐1000 gives us 1000 numbers, one per class, how do we compare them?
- 63. Soqmax • Mul,-‐class version of logis,c func,on • Outputs normalized class “probabili,es” • Takes m inputs and produces m outputs between zero and one, that sum to one • Cross-‐entropy loss • Differen,able
- 65. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 Layer 1
- 66. Learned Filters – Layer1
- 67. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 Layer 2
- 68. Learned Filters – Layer2 Visualizing and Understanding Convolu,onal Networks -‐ Zeiler & Fergus, ECCV 2014
- 69. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 Layer 3
- 70. Learned Filters -‐ Layer3 Visualizing and Understanding Convolu,onal Networks -‐ Zeiler & Fergus, ECCV 2014
- 71. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 Layer 4 Layer 5
- 72. Learned Features – Layers 4 & 5
- 73. Alexnet Architecture -‐ 2012 ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012
- 74. Alexnet Architecture -‐ 2012 Input Conv Relu Pool Conv Relu Pool Conv Relu Conv Relu Conv Relu Pool FC Dropout FC Dropout ImageNet Classifica,on with Deep Convolu,onal Neural Networks Alex Krizhevsky, Ilya Sutskever and Geoffrey E. Hinton Advances in Neural Informa,on Processing Systems 25 eds.F. Pereira, C.J.C. Burges, L. Bo@ou and K.Q. Weinberger pp. 1097-‐1105, 2012 FC 1000
- 77. To Learn More • h@p://colah.github.io/posts/2014-‐07-‐ Understanding-‐Convolu,ons/ • h@ps://adeshpande3.github.io/ adeshpande3.github.io/The-‐9-‐Deep-‐Learning-‐ Papers-‐You-‐Need-‐To-‐Know-‐About.html • h@p://cs231n.github.io/ • h@p://course.fast.ai/
- 78. Ques,ons?