-
-
Notifications
You must be signed in to change notification settings - Fork 0
/
lecture_8.html
267 lines (218 loc) · 11.6 KB
/
lecture_8.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8">
<title>reveal.js – NNML 2020</title>
<meta name="description" content="A framework for easily creating beautiful presentations using HTML">
<meta name="author" content="Hakim El Hattab">
<meta name="apple-mobile-web-app-capable" content="yes">
<meta name="apple-mobile-web-app-status-bar-style" content="black-translucent">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<link rel="stylesheet" href="dist/reset.css">
<link rel="stylesheet" href="dist/reveal.css">
<link rel="stylesheet" href="dist/theme/nmilab.css" id="theme">
<!-- Theme used for syntax highlighting of code -->
<link rel="stylesheet" href="plugin/highlight/monokai.css" id="highlight-theme">
<link rel="stylesheet" type="text/css" href="https://cdn.rawgit.com/dreampulse/computer-modern-web-font/master/fonts.css">
<link rel="stylesheet" type="text/css" href="https://maxcdn.bootstrapcdn.com/font-awesome/4.5.0/css/font-awesome.min.css">
<link rel="stylesheet" type="text/css" href="lib/css/monokai.css">
<link
rel="stylesheet"
href="https://cdnjs.cloudflare.com/ajax/libs/animate.css/4.0.0/animate.min.css"
/>
<link rel="stylesheet" href="reveal.js-plugins/appearance/appearance.css" />
<link rel="stylesheet" href="plugin/highlight/monokai.css" id="highlight-theme" />
<!-- Printing and PDF exports -->
<script>
var link = document.createElement( 'link' );
link.rel = 'stylesheet';
link.type = 'text/css';
link.href = window.location.search.match( /print-pdf/gi ) ? 'css/print/pdf.css' : 'css/print/paper.css';
document.getElementsByTagName( 'head' )[0].appendChild( link );
</script>
<script>
//function for evaluation
function solve()
{
let w0 = document.getElementById("w0").value
let w1 = document.getElementById("w1").value
let b = document.getElementById("b").value
document.getElementById("a11").innerHTML = (eval(w0)*1+eval(w1)*1+eval(b)).toFixed(2)
document.getElementById("a10").innerHTML = (eval(w0)*1+eval(w1)*0+eval(b)).toFixed(2)
document.getElementById("a01").innerHTML = (eval(w0)*0+eval(w1)*1+eval(b)).toFixed(2)
document.getElementById("a00").innerHTML = (eval(w0)*0+eval(w1)*0+eval(b)).toFixed(2)
document.getElementById("y11").innerHTML = (((eval(w0)*1+eval(w1)*1+eval(b))>0)*1)
document.getElementById("y10").innerHTML = (((eval(w0)*1+eval(w1)*0+eval(b))>0)*1)
document.getElementById("y01").innerHTML = (((eval(w0)*0+eval(w1)*1+eval(b))>0)*1)
document.getElementById("y00").innerHTML = (((eval(w0)*0+eval(w1)*0+eval(b))>0)*1)
document.getElementById('w0val').innerHTML=eval(w0);
document.getElementById('w1val').innerHTML=eval(w1);
document.getElementById('bval').innerHTML=eval(b);
}
</script>
<script async defer src="https://buttons.github.io/buttons.js"></script>
</head>
<body>
<div class="reveal">
<!-- Any section element inside of this container is displayed as a slide -->
<div class="slides">
<section data-external="title.html" data-vertical-align-top data-background-color=#B2BA67 ></section>
<section data-markdown data-vertical-align-top data-background-color=#B2BA67><textarea data-template>
<h1> Lecture 8: Generative Adversarial Networks (GANs) <br/> </h1>
</textarea></section>
<section data-markdown><textarea data-template>
<h2> Generative Adversarial Networks - a Brief intro </h2>
<ul>
<li/> A framework for estimating generative models via an adversarial process
<li class=fragment /> Simultaneously train two models:
<ul >
<li/>a generative model G that captures the data distribution
<li />a discriminative model D that estimates the probability that a sample came from the training data rather than G
</ul>
</ul>
<img src="images/GAN_overview.png" class=stretch />
</textarea></section>
<section data-markdown><textarea data-template>
<h2> Generative Adversarial Networks - Examples </h2>
<img src="images/ProgressiveGAN_fig6.png" class=stretch />
<p class=ref><a href=https://arxiv.org/pdf/1710.10196.pdf>Karras et al. 2017</a></p>
</textarea></section>
<section data-markdown><textarea data-template>
<h2> Generative Adversarial Networks - intuition </h2>
<blockquote><p>The generative model can be thought of as analogous to a team of counterfeiters,
trying to produce fake currency and use it without detection, while the discriminative model is analogous to the police, trying to detect the counterfeit currency. Competition in this game drives both teams to improve their methods until the counterfeits are indistiguishable from the genuine articles.</p>
<footer>—Goodfellow, <cite>Generative Adversarial Networks</cite></footer>
</blockquote>
<img src="images/GAN_overview.png" class=stretch />
</textarea></section>
<section data-markdown><textarea data-template>
<h2> Generative Adversarial Networks - Some background </h2>
<ul>
<li/> Zero-sum game
<li/> Minimax
<ul>
<li/> You maximize your payoff
<li/> Your opponent tries to minimize your payoff
</ul>
<li/> Find Nash equilibrium
<ul>
<li/> Each player is assumed to know the equilibrium strategies of the other players, and no player has anything to gain by changing only their own strategy (source: Wikipedia)
</ul>
</ul>
</textarea></section>
<!--
Add general explanation about Generative Modeling itself and bridge the last lecture and GANs
-->
<section data-markdown><textarea data-template>
<h2> Generative Adversarial Networks - Cost and how it works </h2>
$$min_G max_D V(D,G) = \mathbb{E}_{x \sim p_{data}(x)} \left[ logD(x) \right] + \mathbb{E}_{z \sim p_z(z)} \left[ log(1 - D(G(z))) \right]$$
<ul>
<li/> $D$: the probability that a sample came from the training data rather than G
<li class=fragment /> First equation of the right hand side
<ul>
<li/> Sample data $x$ from real distribution
<li/> Since $x$ is sampled from real distribution, $D(x)$ should be close to 1
</ul>
<li class=fragment /> Second equation of the right hand side
<ul>
<li/> Draw a noise $z$ from distribution $P_z$ and generate data $G(z)$
<li/> $G$ want to fool $D$: Want to make D(G(z)) close to 1
</ul>
<ul>
<li/> D wants to maximize (as close to 0 as possible) this quantity. G wants to minimize.
</ul>
</ul>
[](https://colab.research.google.com/drive/1KRvrS_VhPuZPF4Jhglf24E77yfmatcgF?usp=sharing)
</textarea></section>
<section data-markdown data-vertical-align-top><textarea data-template>
<h2> Conditional GAN (cGAN) </h2>
<ul>
<li/> Although GANs showed successful image generation ability, there was no way to control or specify a certain type of image to generate
<li/> Conditional Generative Adversarial Nets was proposed a few months later
</ul>
<img src="images/ConditionalAdversarialNet.png" class=stretch />
</textarea></section>
<section data-markdown data-vertical-align-top><textarea data-template>
<h2> Implement cGAN </h2>
<ul>
<li/> The first cGAN simply concat additional information for conditioning
</ul>
<pre><code class="Python" data-trim data-noescape>
class Generator(nn.Module):
def __init__(self, batch, z_dim, out_shape, num_classes):
super(Generator, self).__init__()
self.batch_size = batch
self.z_dim = z_dim
self.out_shape = out_shape
self.num_classes = num_classes
self.fc1 = nn.Linear(z_dim+num_classes, 256) # simple concat
</code></pre>
</textarea></section>
</div>
</div>
<script src="dist/reveal.js"></script>
<script src="plugin/zoom/zoom.js"></script>
<script src="plugin/notes/notes.js"></script>
<script src="plugin/math/math.js"></script>
<script src="plugin/search/search.js"></script>
<script src="plugin/markdown/markdown.js"></script>
<script src="plugin/highlight/highlight.js"></script>
<script src="reveal.js-plugins/menu/menu.js"></script>
<script src="reveal.js-plugins/audio-slideshow/plugin.js"></script>
<script src="reveal.js-plugins/audio-slideshow/recorder.js"></script>
<script src="reveal.js-plugins/audio-slideshow/RecordRTC.js"></script>
<script src="reveal.js-plugins/chalkboard/plugin.js"></script>
<script src="reveal.js-plugins/external/external.js"></script>
<script src="reveal.js-plugins/externalcode/externalcode.js"></script>
<script src="reveal.js-plugins/appearance/appearance.js"></script>
<script>
Reveal.configure({ pdfMaxPagesPerSlide: 1, hash: true, slideNumber: true})
// Also available as an ES module, see:
// https://revealjs.com/initialization/
Reveal.initialize({
mouseWheel: false,
width: 1280,
height: 720,
margin: 0.0,
navigationMode: 'grid',
transition: 'fade',
controls: true,
progress: true,
center: true,
hash: true,
// Learn about plugins: https://revealjs.com/plugins/
plugins: [Appearance, RevealZoom, RevealSearch, RevealMarkdown, RevealHighlight, RevealAudioRecorder, RevealAudioSlideshow, RevealMath, RevealExternal, Externalcode, RevealMenu],
menu: { // Menu works best with font-awesome installed: sudo apt-get install fonts-font-awesome
themes: false,
transitions: false,
markers: true,
hideMissingTitles: true,
custom: [
{ title: 'Plugins', icon: '<i class="fa fa-external-link-alt"></i>', src: 'toc.html' },
{ title: 'About', icon: '<i class="fa fa-info"></i>', src: 'about.html' }
]
},
audio: {
prefix: 'audio/lecture_8/', // audio files are stored in the "audio" folder
suffix: '.ogg', // audio files have the ".ogg" ending
textToSpeechURL: null, // the URL to the text to speech converter
defaultNotes: false, // use slide notes as default for the text to speech converter
defaultText: false, // use slide text as default for the text to speech converter
advance: 300, // advance to next slide after given time in milliseconds after audio has played, use negative value to not advance
autoplay: false, // automatically start slideshow
defaultDuration: 1.0, // default duration in seconds if no audio is available
defaultAudios: true, // try to play audios with names such as audio/1.2.ogg
playerOpacity: 0.5, // opacity value of audio player if unfocused
playerStyle: 'position: fixed; bottom: 4px; left: 25%; width: 50%; height:75px; z-index: 33;', // style used for container of audio controls
startAtFragment: false, // when moving to a slide, start at the current fragment or at the start of the slide
},
// ...
keyboard: {
82: function() { Recorder.toggleRecording(); }, // press 'r' to start/stop recording
90: function() { Recorder.downloadZip(); }, // press 'z' to download zip containing audio files
84: function() { Recorder.fetchTTS(); } // press 't' to fetch TTS audio files
}
});
</script>
</body>
</html>