microsoft/AI-For-Beginners

Public

mirrored fromhttps://github.com/microsoft/AI-For-BeginnersAvailable

CodeCommitsIssuesPull requestsActionsInsightsSecurity
9b66995013b40a5c2d7673e7672fac58beac6085

Branches

Tags

  • No tags available.
0Branches0Tags
Go to file
Add file
Code

Clone

HTTPS

Download ZIP

lessons/4-ComputerVision/07-ConvNets/ConvNetsTF.ipynb

674lines · modecode

1{
2 "cells": [
3 {
4 "cell_type": "markdown",
5 "metadata": {},
6 "source": [
7 "# Convolutional neural networks\n",
8 "\n",
9 "We have seen before that neural networks are quite good at dealing with images, and even one-layer perceptron is able to recognize handwritten digits from MNIST dataset with reasonable accuracy. However, MNIST dataset is very special, and all digits are centered inside the image, which makes the task simpler.\n",
10 "\n",
11 "In real life, we want to be able to recognize objects on the picture regardless of their exact location in the image. Computer vision is different from generic classification, because when we are trying to find a certain object in the picture, we are scanning the image looking for some specific **patterns** and their combinations. For example, when looking for a cat, we first may look for horizontal lines, which can form whiskers, and then certain combination of whiskers can tell us that it is actually a picture of a cat. Relative position and presence of certain patterns is important, and not their exact position on the image. \n",
12 "\n",
13 "To extract patterns, we will use the notion of **convolutional filters**. But first, let us load all dependencies and functions that we have defined in the previous units. We will also import `tfcv` helper library that contain some useful functions that we do not want to define inside this notebook to keep the code short and clean. "
14 ]
15 },
16 {
17 "cell_type": "code",
18 "execution_count": 1,
19 "metadata": {
20 "tags": []
21 },
22 "outputs": [],
23 "source": [
24 "import tensorflow as tf\n",
25 "from tensorflow import keras\n",
26 "import matplotlib.pyplot as plt\n",
27 "import numpy as np\n",
28 "from tfcv import *"
29 ]
30 },
31 {
32 "cell_type": "markdown",
33 "metadata": {},
34 "source": [
35 "In this example, we will focus on the MNIST dataset that we have seen before, and on image classification. We will start by loading the dataset using Keras built-in functions."
36 ]
37 },
38 {
39 "cell_type": "code",
40 "execution_count": 2,
41 "metadata": {},
42 "outputs": [],
43 "source": [
44 "(x_train,y_train),(x_test,y_test) = keras.datasets.mnist.load_data()\n",
45 "x_train = x_train.astype(np.float32) / 255.0\n",
46 "x_test = x_test.astype(np.float32) / 255.0"
47 ]
48 },
49 {
50 "cell_type": "markdown",
51 "metadata": {},
52 "source": [
53 "## Convolutional filters\n",
54 "\n",
55 "Convolutional filters are small windows that run over each pixel of the image and compute weighted average of the neighboring pixels.\n",
56 "\n",
57 "\n",
58 "\n",
59 "They are defined by matrices of weight coefficients. Let's see the examples of applying two different convolutional filters over our MNIST handwritten digits:"
60 ]
61 },
62 {
63 "cell_type": "code",
64 "execution_count": 3,
65 "metadata": {},
66 "outputs": [
67 {
68 "data": {
69 "image/png": "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",
70 "text/plain": [
71 "<Figure size 576x216 with 12 Axes>"
72 ]
73 },
74 "metadata": {
75 "needs_background": "light"
76 },
77 "output_type": "display_data"
78 },
79 {
80 "data": {
81 "image/png": "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",
82 "text/plain": [
83 "<Figure size 576x216 with 12 Axes>"
84 ]
85 },
86 "metadata": {
87 "needs_background": "light"
88 },
89 "output_type": "display_data"
90 }
91 ],
92 "source": [
93 "plot_convolution(x_train[:5],[[-1.,0.,1.],[-1.,0.,1.],[-1.,0.,1.]],'Vertical edge filter')\n",
94 "plot_convolution(x_train[:5],[[-1.,-1.,-1.],[0.,0.,0.],[1.,1.,1.]],'Horizontal edge filter')"
95 ]
96 },
97 {
98 "cell_type": "markdown",
99 "metadata": {},
100 "source": [
101 "First filter is called a **vertical edge filter**, and it is defined by the following matrix:\n",
102 "$$\n",
103 "\\left(\n",
104 " \\begin{matrix}\n",
105 " -1 & 0 & 1 \\cr\n",
106 " -1 & 0 & 1 \\cr\n",
107 " -1 & 0 & 1 \\cr\n",
108 " \\end{matrix}\n",
109 "\\right)\n",
110 "$$\n",
111 "When this filter goes over relatively uniform pixel field, all values add up to 0. However, when it encounters a vertical edge in the image, high spike value is generated. That's why in the images above you can see vertical edges represented by high and low values, while horizontal edges are averaged out.\n",
112 "\n",
113 "An opposite thing happens when we apply horizontal edge filter - horizontal lines are amplified, and vertical are averaged out.\n",
114 "\n",
115 "In classical computer vision, multiple filters were applied to the image to generate features, which then were used by machine learning algorithm to build a classifier. Those filters are in fact similar to neural structures that are available in the vision system of some animals.\n",
116 "\n",
117 "<img src=\"images/lmfilters.jpg\" width=\"400\"/>\n",
118 "\n",
119 "However, in deep learning we construct networks that **learn** best convolutional filters to solve classification problem. To do that, we introduce **convolutional layers**."
120 ]
121 },
122 {
123 "cell_type": "markdown",
124 "metadata": {},
125 "source": [
126 "## Covolutional layers\n",
127 "\n",
128 "To make the weights of convolutional layer trainable, we need somehow to reduce the process of applying convolutional filter window to the image to the matrix operations, which can then be subject to backward propagation training. To do this, we use a clever matrix transformation, which we call **im2col**.\n",
129 "\n",
130 "Suppose we have a small image $\\mathbf{x}$, with the following pixels:\n",
131 "\n",
132 "$$\n",
133 "\\mathbf{x} = \\left(\n",
134 " \\begin{array}{ccccc}\n",
135 " a & b & c & d & e \\\\\n",
136 " f & g & h & i & j \\\\\n",
137 " k & l & m & n & o \\\\\n",
138 " p & q & r & s & t \\\\\n",
139 " u & v & w & x & y \\\\\n",
140 " \\end{array}\n",
141 " \\right)\n",
142 "$$\n",
143 "\n",
144 "And we want to apply two conv filters, with the following weights:\n",
145 "$$\n",
146 "W^{(i)} = \\left(\\begin{array}{ccc}\n",
147 " w^{(i)}_{00} & w^{(i)}_{01} & w^{(i)}_{02} \\\\\n",
148 " w^{(i)}_{10} & w^{(i)}_{11} & w^{(i)}_{12} \\\\\n",
149 " w^{(i)}_{20} & w^{(i)}_{21} & w^{(i)}_{22} \\\\\n",
150 " \\end{array}\\right) \n",
151 "$$\n",
152 "\n",
153 "When applying the convolution, the first pixel of the result would be obtained by element-wise multiplication of \n",
154 "$\\left(\\begin{array}{ccc}\n",
155 " a & b & c \\\\\n",
156 " f & g & h \\\\\n",
157 " k & l & m \\\\\n",
158 "\\end{array}\\right)$ and $W^{(i)}$, the second element - by multiplying by $\\left(\\begin{array}{ccc}\n",
159 " b & c & d \\\\\n",
160 " g & h & i \\\\\n",
161 " l & m & n \\\\\n",
162 "\\end{array}\\right)$ by $W^{(i)}$, and so on.\n",
163 "\n",
164 "To formalize this process, let's extract all $3\\times3$ fragments of the original image $x$ into the following matrix:\n",
165 "\n",
166 "$$\n",
167 "\\mathrm{im2col}(x) = \\left[\n",
168 " \\begin{array}{cccccc}\n",
169 " a & b & \\ldots & g & \\ldots & m \\\\\n",
170 " b & c & \\ldots & h & \\ldots & n \\\\\n",
171 " c & d & \\ldots & i & \\ldots & o \\\\\n",
172 " f & g & \\ldots & l & \\ldots & r \\\\\n",
173 " g & h & \\ldots & m & \\ldots & s \\\\\n",
174 " h & i & \\ldots & n & \\ldots & t \\\\\n",
175 " k & l & \\ldots & q & \\ldots & w \\\\\n",
176 " l & m & \\ldots & r & \\ldots & x \\\\\n",
177 " m & n & \\ldots & s & \\ldots & y \\\\\n",
178 " \\end{array}\n",
179 " \\right]\n",
180 "$$\n",
181 "\n",
182 "Each column of this matrix corresponds to each $3\\times3$ subregion of the original image. Now, to get the result of the convolution, we just need to multiply this matrix by the matrix or weights\n",
183 "$$\n",
184 "\\mathbf{W} = \\left[\n",
185 " \\begin{array}{cccccccc}\n",
186 " w^{(0)}_{00} & w^{(0)}_{01} & w^{(0)}_{02} & w^{(0)}_{10} & w^{(0)}_{11} & \\ldots & w^{(0)}_{21} & w^{(0)}_{22} \\\\\n",
187 " w^{(1)}_{00} & w^{(1)}_{01} & w^{(1)}_{02} & w^{(1)}_{10} & w^{(1)}_{11} & \\ldots & w^{(1)}_{21} & w^{(1)}_{22} \\\\\n",
188 " \\end{array}\n",
189 " \\right]\n",
190 "$$\n",
191 "(each row of this matrix contains weights of $i$-th filter, flattened into one row)\n",
192 "\n",
193 "So the application of a convolution filter to the original image can be replaced by matrix multiplication, which we already know how to handle using back prop:\n",
194 "$$\n",
195 "C(x) = W\\times\\mathbf{im2col}(x)\n",
196 "$$\n",
197 "\n",
198 "Convolutional layers are defined using `Conv2d` class. We need to specify the following:\n",
199 "* `filters` - number of filters to use. We will use 9 different filters, which will give the network plenty of opportunities to explore which filters work best for our scenario.\n",
200 "* `kernel_size` is the size of the sliding window. Usually 3x3 or 5x5 filters are used.\n",
201 "\n",
202 "Simplest CNN will contain one convolutional layer. Given the input size 28x28, after applying nine 5x5 filters we will end up with a tensor of 24x24x9. The spatial dimension is smaller, because there are only 24 positions where a sliding interval of length 5 can fit into 28 pixels).\n",
203 "\n",
204 "After convolution, we flatten 24x24x9 tensor into one vector of size 5184, and then add linear layer, to produce 10 classes. We also use `relu` activation function in between layers. "
205 ]
206 },
207 {
208 "cell_type": "code",
209 "execution_count": 4,
210 "metadata": {},
211 "outputs": [
212 {
213 "name": "stdout",
214 "output_type": "stream",
215 "text": [
216 "Model: \"sequential\"\n",
217 "_________________________________________________________________\n",
218 " Layer (type) Output Shape Param # \n",
219 "=================================================================\n",
220 " conv2d (Conv2D) (None, 24, 24, 9) 234 \n",
221 " \n",
222 " flatten (Flatten) (None, 5184) 0 \n",
223 " \n",
224 " dense (Dense) (None, 10) 51850 \n",
225 " \n",
226 "=================================================================\n",
227 "Total params: 52,084\n",
228 "Trainable params: 52,084\n",
229 "Non-trainable params: 0\n",
230 "_________________________________________________________________\n"
231 ]
232 }
233 ],
234 "source": [
235 "model = keras.models.Sequential([\n",
236 " keras.layers.Conv2D(filters=9, kernel_size=(5,5), input_shape=(28,28,1),activation='relu'),\n",
237 " keras.layers.Flatten(),\n",
238 " keras.layers.Dense(10)\n",
239 "])\n",
240 "\n",
241 "model.compile(loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),metrics=['acc'])\n",
242 "\n",
243 "model.summary()"
244 ]
245 },
246 {
247 "cell_type": "markdown",
248 "metadata": {},
249 "source": [
250 "You can see that this network contains around 50k trainable parameters, compared to around 80k in fully-connected multi-layered networks. This allows us to achieve good results even on smaller datasets, because convolutional networks generalize much better.\n",
251 "\n",
252 "> **Note**: In most of the practical cases, we want to apply convolutional layers to color images. Thus, `Conv2D` layer expects the input to be of the shape $W\\times H\\times C$, where $W$ and $H$ are width and height of the image, and $C$ is the number of color channels. For grayscale images, we need the same shape with $C=1$.\n",
253 "\n",
254 "We need to reshape our data before starting training:"
255 ]
256 },
257 {
258 "cell_type": "code",
259 "execution_count": 5,
260 "metadata": {},
261 "outputs": [
262 {
263 "name": "stdout",
264 "output_type": "stream",
265 "text": [
266 "Epoch 1/5\n",
267 "1875/1875 [==============================] - 15s 7ms/step - loss: 0.2099 - acc: 0.9410 - val_loss: 0.0879 - val_acc: 0.9735\n",
268 "Epoch 2/5\n",
269 "1875/1875 [==============================] - 13s 7ms/step - loss: 0.0858 - acc: 0.9753 - val_loss: 0.0682 - val_acc: 0.9791\n",
270 "Epoch 3/5\n",
271 "1875/1875 [==============================] - 13s 7ms/step - loss: 0.0665 - acc: 0.9808 - val_loss: 0.0553 - val_acc: 0.9829\n",
272 "Epoch 4/5\n",
273 "1875/1875 [==============================] - 15s 8ms/step - loss: 0.0582 - acc: 0.9835 - val_loss: 0.0513 - val_acc: 0.9835\n",
274 "Epoch 5/5\n",
275 "1875/1875 [==============================] - 14s 8ms/step - loss: 0.0527 - acc: 0.9847 - val_loss: 0.0503 - val_acc: 0.9833\n"
276 ]
277 }
278 ],
279 "source": [
280 "x_train_c = np.expand_dims(x_train,3)\n",
281 "x_test_c = np.expand_dims(x_test,3)\n",
282 "hist = model.fit(x_train_c,y_train,validation_data=(x_test_c,y_test),epochs=5)"
283 ]
284 },
285 {
286 "cell_type": "code",
287 "execution_count": null,
288 "metadata": {},
289 "outputs": [
290 {
291 "data": {
292 "image/png": "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",
293 "text/plain": [
294 "<Figure size 1080x216 with 2 Axes>"
295 ]
296 },
297 "metadata": {
298 "needs_background": "light"
299 },
300 "output_type": "display_data"
301 }
302 ],
303 "source": [
304 "plot_results(hist)"
305 ]
306 },
307 {
308 "cell_type": "markdown",
309 "metadata": {},
310 "source": [
311 "As you can see, we are able to achieve higher accuracy, and much faster (in terms of number of epochs), compared to the fully-connected networks from previous unit. However, the training itself requires more resources, and may be slower on non-GPU computers.\n",
312 "\n",
313 "## Visualizing Convolutional Layers\n",
314 "\n",
315 "We can also visualize the weights of our trained convolutional layers, to try and make some more sense of what is going on:"
316 ]
317 },
318 {
319 "cell_type": "code",
320 "execution_count": null,
321 "metadata": {},
322 "outputs": [
323 {
324 "data": {
325 "image/png": "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",
326 "text/plain": [
327 "<Figure size 432x288 with 9 Axes>"
328 ]
329 },
330 "metadata": {
331 "needs_background": "light"
332 },
333 "output_type": "display_data"
334 }
335 ],
336 "source": [
337 "fig,ax = plt.subplots(1,9)\n",
338 "l = model.layers[0].weights[0]\n",
339 "for i in range(9):\n",
340 " ax[i].imshow(l[...,0,i])\n",
341 " ax[i].axis('off')"
342 ]
343 },
344 {
345 "cell_type": "markdown",
346 "metadata": {},
347 "source": [
348 "You can see that some of those filters look like they can recognize some oblique strokes, while others look pretty random. \n",
349 "\n",
350 "> **Task**: Train the same network with 3x3 filters and visualize them. Do you see more familiar patterns?\n",
351 "\n",
352 "## Multi-layered CNNs and pooling layers\n",
353 "\n",
354 "First convolutional layers looks for primitive patterns, such as horizontal or vertical lines, but we can apply further convolutional layers on top of them to look for higher-level patterns, such as primitive shapes. Then more convolutional layers can combine those shapes into some parts of the picture, up to the final object that we are trying to classify. \n",
355 "\n",
356 "When doing so, we may also apply one trick: reducing the spatial size of the image. Once we have detected there is a horizontal stoke within sliding 3x3 window, it is not so important at which exact pixel it occurred. Thus we can \"scale down\" the size of the image, which is done using one of the **pooling layers**:\n",
357 "\n",
358 " * **Average Pooling** takes a sliding window (for example, 2x2 pixels) and computes an average of values within the window\n",
359 " * **Max Pooling** replaces the window with the maximum value. The idea behind max pooling is to detect a presence of a certain pattern within the sliding window.\n",
360 "\n",
361 "Thus, in a typical CNN there would be several convolutional layers, with pooling layers in between them to decrease dimensions of the image. We would also increase the number of filters, because as patterns become more advanced - there are more possible interesting combinations that we need to be looking for.\n",
362 "\n",
363 "![An image showing several convolutional layers with pooling layers.](./images/cnn-pyramid.png)\n",
364 "\n",
365 "Because of decreasing spatial dimensions and increasing feature/filters dimensions, this architecture is also called **pyramid architecture**. "
366 ]
367 },
368 {
369 "cell_type": "code",
370 "execution_count": null,
371 "metadata": {},
372 "outputs": [
373 {
374 "name": "stdout",
375 "output_type": "stream",
376 "text": [
377 "Model: \"sequential_1\"\n",
378 "_________________________________________________________________\n",
379 "Layer (type) Output Shape Param # \n",
380 "=================================================================\n",
381 "conv2d_1 (Conv2D) (None, 24, 24, 10) 260 \n",
382 "_________________________________________________________________\n",
383 "max_pooling2d (MaxPooling2D) (None, 12, 12, 10) 0 \n",
384 "_________________________________________________________________\n",
385 "conv2d_2 (Conv2D) (None, 8, 8, 20) 5020 \n",
386 "_________________________________________________________________\n",
387 "max_pooling2d_1 (MaxPooling2 (None, 4, 4, 20) 0 \n",
388 "_________________________________________________________________\n",
389 "flatten_1 (Flatten) (None, 320) 0 \n",
390 "_________________________________________________________________\n",
391 "dense_1 (Dense) (None, 10) 3210 \n",
392 "=================================================================\n",
393 "Total params: 8,490\n",
394 "Trainable params: 8,490\n",
395 "Non-trainable params: 0\n",
396 "_________________________________________________________________\n"
397 ]
398 }
399 ],
400 "source": [
401 "model = keras.models.Sequential([\n",
402 " keras.layers.Conv2D(filters=10, kernel_size=(5,5), input_shape=(28,28,1),activation='relu'),\n",
403 " keras.layers.MaxPooling2D(),\n",
404 " keras.layers.Conv2D(filters=20, kernel_size=(5,5), activation='relu'),\n",
405 " keras.layers.MaxPooling2D(), \n",
406 " keras.layers.Flatten(),\n",
407 " keras.layers.Dense(10)\n",
408 "])\n",
409 "\n",
410 "model.compile(loss=keras.losses.SparseCategoricalCrossentropy(from_logits=True),metrics=['acc'])\n",
411 "\n",
412 "model.summary()"
413 ]
414 },
415 {
416 "cell_type": "markdown",
417 "metadata": {},
418 "source": [
419 "Notice that the number of trainable parameters (~8.5K) is dramatically smaller than in previous cases. This happens because convolutional layers in general have few parameters, and dimensionality of the image before applying final dense layer is significantly reduced. Small number of parameters have positive impact on our models, because it helps to prevent overfitting even on smaller dataset sizes."
420 ]
421 },
422 {
423 "cell_type": "code",
424 "execution_count": null,
425 "metadata": {},
426 "outputs": [
427 {
428 "name": "stdout",
429 "output_type": "stream",
430 "text": [
431 "Epoch 1/5\n",
432 "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0723 - acc: 0.9780 - val_loss: 0.0423 - val_acc: 0.9861\n",
433 "Epoch 2/5\n",
434 "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0523 - acc: 0.9842 - val_loss: 0.0425 - val_acc: 0.9866\n",
435 "Epoch 3/5\n",
436 "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0448 - acc: 0.9868 - val_loss: 0.0403 - val_acc: 0.9865\n",
437 "Epoch 4/5\n",
438 "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0383 - acc: 0.9886 - val_loss: 0.0323 - val_acc: 0.9888\n",
439 "Epoch 5/5\n",
440 "1875/1875 [==============================] - 6s 3ms/step - loss: 0.0338 - acc: 0.9895 - val_loss: 0.0331 - val_acc: 0.9896\n"
441 ]
442 }
443 ],
444 "source": [
445 "hist = model.fit(x_train_c,y_train,validation_data=(x_test_c,y_test),epochs=5)"
446 ]
447 },
448 {
449 "cell_type": "code",
450 "execution_count": null,
451 "metadata": {},
452 "outputs": [
453 {
454 "data": {
455 "image/png": "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",
456 "text/plain": [
457 "<Figure size 1080x216 with 2 Axes>"
458 ]
459 },
460 "metadata": {
461 "needs_background": "light"
462 },
463 "output_type": "display_data"
464 }
465 ],
466 "source": [
467 "plot_results(hist)"
468 ]
469 },
470 {
471 "cell_type": "markdown",
472 "metadata": {},
473 "source": [
474 "What you should probably observe is that we are able to achieve higher accuracy than with just one layer, and much faster in terms of number of epochs - just with 1 or 2 epochs. It means that sophisticated network architecture needs much fewer data to figure out what is going on, and to extract generic patterns from our images. However, training also takes longer, and requires a GPU.\n",
475 "\n",
476 "## Playing with real images from the CIFAR-10 dataset\n",
477 "\n",
478 "While our handwritten digit recognition problem may seem like a toy problem, we are now ready to do something more serious. Let's explore more advanced dataset of pictures of different objects, called [CIFAR-10](https://www.cs.toronto.edu/~kriz/cifar.html). It contains 60k 32x32 images, divided into 10 classes. "
479 ]
480 },
481 {
482 "cell_type": "code",
483 "execution_count": null,
484 "metadata": {},
485 "outputs": [],
486 "source": [
487 "(x_train,y_train),(x_test,y_test) = keras.datasets.cifar10.load_data()\n",
488 "x_train = x_train.astype(np.float32) / 255.0\n",
489 "x_test = x_test.astype(np.float32) / 255.0\n",
490 "classes = ('plane', 'car', 'bird', 'cat',\n",
491 " 'deer', 'dog', 'frog', 'horse', 'ship', 'truck')"
492 ]
493 },
494 {
495 "cell_type": "code",
496 "execution_count": null,
497 "metadata": {},
498 "outputs": [
499 {
500 "data": {
501 "image/png": "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",
502 "text/plain": [
503 "<Figure size 1080x216 with 10 Axes>"
504 ]
505 },
506 "metadata": {
507 "needs_background": "light"
508 },
509 "output_type": "display_data"
510 }
511 ],
512 "source": [
513 "display_dataset(x_train,y_train,classes=classes)"
514 ]
515 },
516 {
517 "cell_type": "markdown",
518 "metadata": {},
519 "source": [
520 "A well-known architecture for CIFAR-10 is called [LeNet](https://en.wikipedia.org/wiki/LeNet), and has been proposed by *Yann LeCun*. It follows the same principles as we have outlined above, the main difference being 3 input color channels instead of 1. "
521 ]
522 },
523 {
524 "cell_type": "code",
525 "execution_count": null,
526 "metadata": {},
527 "outputs": [
528 {
529 "name": "stdout",
530 "output_type": "stream",
531 "text": [
532 "Model: \"sequential_3\"\n",
533 "_________________________________________________________________\n",
534 "Layer (type) Output Shape Param # \n",
535 "=================================================================\n",
536 "conv2d_8 (Conv2D) (None, 28, 28, 6) 456 \n",
537 "_________________________________________________________________\n",
538 "max_pooling2d_6 (MaxPooling2 (None, 14, 14, 6) 0 \n",
539 "_________________________________________________________________\n",
540 "conv2d_9 (Conv2D) (None, 10, 10, 16) 2416 \n",
541 "_________________________________________________________________\n",
542 "max_pooling2d_7 (MaxPooling2 (None, 5, 5, 16) 0 \n",
543 "_________________________________________________________________\n",
544 "flatten_3 (Flatten) (None, 400) 0 \n",
545 "_________________________________________________________________\n",
546 "dense_9 (Dense) (None, 120) 48120 \n",
547 "_________________________________________________________________\n",
548 "dense_10 (Dense) (None, 84) 10164 \n",
549 "_________________________________________________________________\n",
550 "dense_11 (Dense) (None, 10) 850 \n",
551 "=================================================================\n",
552 "Total params: 62,006\n",
553 "Trainable params: 62,006\n",
554 "Non-trainable params: 0\n",
555 "_________________________________________________________________\n"
556 ]
557 }
558 ],
559 "source": [
560 "model = keras.models.Sequential([\n",
561 " keras.layers.Conv2D(filters = 6, kernel_size = 5, strides = 1, activation = 'relu', input_shape = (32,32,3)),\n",
562 " keras.layers.MaxPooling2D(pool_size = 2, strides = 2),\n",
563 " keras.layers.Conv2D(filters = 16, kernel_size = 5, strides = 1, activation = 'relu'),\n",
564 " keras.layers.MaxPooling2D(pool_size = 2, strides = 2),\n",
565 " keras.layers.Flatten(),\n",
566 " keras.layers.Dense(120, activation = 'relu'),\n",
567 " keras.layers.Dense(84, activation = 'relu'),\n",
568 " keras.layers.Dense(10, activation = 'softmax')])\n",
569 "\n",
570 "model.summary()"
571 ]
572 },
573 {
574 "cell_type": "markdown",
575 "metadata": {},
576 "source": [
577 "Training this network properly will take significant amount of time, and should preferably be done on GPU-enabled compute."
578 ]
579 },
580 {
581 "cell_type": "code",
582 "execution_count": null,
583 "metadata": {},
584 "outputs": [
585 {
586 "name": "stdout",
587 "output_type": "stream",
588 "text": [
589 "Epoch 1/10\n",
590 "1563/1563 [==============================] - 6s 4ms/step - loss: 1.6205 - acc: 0.4033 - val_loss: 1.4287 - val_acc: 0.4743\n",
591 "Epoch 2/10\n",
592 "1563/1563 [==============================] - 6s 4ms/step - loss: 1.3435 - acc: 0.5154 - val_loss: 1.2804 - val_acc: 0.5412\n",
593 "Epoch 3/10\n",
594 "1563/1563 [==============================] - 5s 3ms/step - loss: 1.2225 - acc: 0.5645 - val_loss: 1.2164 - val_acc: 0.5600\n",
595 "Epoch 4/10\n",
596 "1563/1563 [==============================] - 5s 4ms/step - loss: 1.1360 - acc: 0.5957 - val_loss: 1.1918 - val_acc: 0.5768\n",
597 "Epoch 5/10\n",
598 "1563/1563 [==============================] - 5s 3ms/step - loss: 1.0776 - acc: 0.6178 - val_loss: 1.1451 - val_acc: 0.5906\n",
599 "Epoch 6/10\n",
600 "1563/1563 [==============================] - 5s 4ms/step - loss: 1.0228 - acc: 0.6370 - val_loss: 1.1178 - val_acc: 0.6098\n",
601 "Epoch 7/10\n",
602 "1563/1563 [==============================] - 5s 4ms/step - loss: 0.9769 - acc: 0.6544 - val_loss: 1.0793 - val_acc: 0.6202\n",
603 "Epoch 8/10\n",
604 "1563/1563 [==============================] - 5s 3ms/step - loss: 0.9340 - acc: 0.6711 - val_loss: 1.0783 - val_acc: 0.6271\n",
605 "Epoch 9/10\n",
606 "1563/1563 [==============================] - 5s 4ms/step - loss: 0.8983 - acc: 0.6824 - val_loss: 1.0952 - val_acc: 0.6203\n",
607 "Epoch 10/10\n",
608 "1563/1563 [==============================] - 6s 4ms/step - loss: 0.8648 - acc: 0.6939 - val_loss: 1.1103 - val_acc: 0.6217\n"
609 ]
610 }
611 ],
612 "source": [
613 "model.compile(optimizer = 'adam', loss = 'sparse_categorical_crossentropy', metrics = ['acc'])\n",
614 "hist = model.fit(x_train,y_train,validation_data=(x_test,y_test),epochs=10)"
615 ]
616 },
617 {
618 "cell_type": "code",
619 "execution_count": null,
620 "metadata": {},
621 "outputs": [
622 {
623 "data": {
624 "image/png": "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",
625 "text/plain": [
626 "<Figure size 1080x216 with 2 Axes>"
627 ]
628 },
629 "metadata": {
630 "needs_background": "light"
631 },
632 "output_type": "display_data"
633 }
634 ],
635 "source": [
636 "plot_results(hist)"
637 ]
638 },
639 {
640 "cell_type": "markdown",
641 "metadata": {},
642 "source": [
643 "The accuracy that we have been able to achieve with few epochs of training does not seem too great. However, remember that bling guessing would only give us 10% accuracy, and that our problem is actually significantly more difficult than MNIST digit classification. Getting above 50% accuracy in such a short training time seems like a good accomplishment.\n",
644 "\n",
645 "## Takeaways\n",
646 "\n",
647 "In this unit, we have learned the main concept behind computer vision neural networks - convolutional networks. Real-life architectures that power image classification, object detection, and even image generation networks are all based on CNNs, just with more layers and some additional training tricks."
648 ]
649 }
650 ],
651 "metadata": {
652 "interpreter": {
653 "hash": "0cb620c6d4b9f7a635928804c26cf22403d89d98d79684e4529119355ee6d5a5"
654 },
655 "kernelspec": {
656 "display_name": "Python 3.8.12 64-bit (conda)",
657 "name": "python3"
658 },
659 "language_info": {
660 "codemirror_mode": {
661 "name": "ipython",
662 "version": 3
663 },
664 "file_extension": ".py",
665 "mimetype": "text/x-python",
666 "name": "python",
667 "nbconvert_exporter": "python",
668 "pygments_lexer": "ipython3",
669 "version": "3.8.12"
670 }
671 },
672 "nbformat": 4,
673 "nbformat_minor": 2
674}