Download - DL Classe 2 - I See what you mean
![Page 1: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/1.jpg)
Louis Monier@louis_monier
https://www.linkedin.com/in/louismonier
Deep Learning for ImagesI see what you mean...
Gregory Renard@redohttps://www.linkedin.com/in/gregoryrenard
Class 2 - Q2 - 2016
![Page 2: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/2.jpg)
Fun with Images
Image Classification: kayak, boy Entity Detection
kayak boy
Face Recognition
LeoGollum
![Page 3: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/3.jpg)
More Fun with Images
Pose DetectionImage SegmentationImage Captioning: “A young boy wearing an orange vest riding a yellow kayak on water, with sunlight reflections.”
![Page 4: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/4.jpg)
Yet more Fun with Images
Optical Character Recognition (OCR): Astronomy is the science which treats of the nature and properties of the heavenly bodies.
Autonomous Vehicles
Handwriting Recognition: combustible: “able to catch fire”, adjective for being capable of igniting and burning.
![Page 5: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/5.jpg)
Our Wet Hardware
Alternating layers of- simple cells (filters)- complex cells (combination)
Simple patterns to abstract concepts
~ 5B neurons for vision
![Page 6: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/6.jpg)
Convolutional Neural Network (ConvNet, CNN)Suggested by Kunihiko Fukushima, 1980
LeNet, by Yann LeCun, 1998, to classify hand-written digits
![Page 7: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/7.jpg)
filterimage
= 6.6
= -7.8
1.0 - really want
0.2 - sort of want
-1.0 - don’t want
Convolution: Applying a Filter to a Signal
through
through
=
=
image filter
1.0
0.5
0.0
through
through
![Page 8: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/8.jpg)
Convolutional Layer - Basic Unit
5x5x3 chunk of inputs
Layer N Layer N+1
ReLU neuron
(3)
(3)
(3)
(3)
5 x 5 x 3 = 75 inputs76 weights
![Page 9: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/9.jpg)
Convolutional Layer: Add Depth
5x5x3 chunk of inputs
Layer N Layer N+1
Depth = 7 ReLU neurons in parallel,with different weights
![Page 10: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/10.jpg)
Stride = 1
Convolutional Layer: Repeat over entire image
L=W=5
D=4
zero padding
D=7
Shared weights!!!
![Page 11: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/11.jpg)
Pooling Layer: Squeeeeeze!
Max Pooling Average Pooling
Layer N+1Layer N
![Page 12: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/12.jpg)
Classical CNN topology - VGGNet (2013)
224x224 112x112 56x56 28x28 14x14 FC
D=64
D=128
D=256
D=512
D=512
D=4096 D=4096 D=1000
FC FC + Softmax
ConvNetPool
![Page 13: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/13.jpg)
Layer 1
Filter Matching images
![Page 14: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/14.jpg)
Layer 2
![Page 15: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/15.jpg)
Layer 3
![Page 16: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/16.jpg)
Layer 4
![Page 17: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/17.jpg)
Layer 5
![Page 18: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/18.jpg)
Modern ConvNet - GoogLeNet
GoogLeNet (2014)
ResNet-34 (2015)
![Page 19: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/19.jpg)
Manifolds
![Page 20: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/20.jpg)
Real-life Data vs Random Data
If music be the food of love, play on!-- William Shakespeare
3Flr'kI5;LS3oLj1xK52,BA1 Rea5IYSf-- 1000 monkeys typing
-- Real world -- Random Pixels
![Page 21: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/21.jpg)
![Page 22: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/22.jpg)
Workshop : Keras & MNISThttps://github.com/holbertonschool/deep-learning/tree/master/Class%20%232
![Page 23: DL Classe 2 - I See what you mean](https://reader033.vdocuments.us/reader033/viewer/2022052606/58f2cb741a28abb15c8b45d9/html5/thumbnails/23.jpg)
Workshop : Keras & CIFAR 10https://github.com/holbertonschool/deep-learning/tree/master/Class%20%232