-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
225 lines (188 loc) · 11.9 KB
/
index.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
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<title>Darkon Documentation — darkon 0.0.6 documentation</title>
<link rel="stylesheet" href="_static/alabaster.css" type="text/css" />
<link rel="stylesheet" href="_static/pygments.css" type="text/css" />
<script type="text/javascript">
var DOCUMENTATION_OPTIONS = {
URL_ROOT: './',
VERSION: '0.0.6',
COLLAPSE_INDEX: false,
FILE_SUFFIX: '.html',
HAS_SOURCE: true,
SOURCELINK_SUFFIX: '.txt'
};
</script>
<script type="text/javascript" src="_static/jquery.js"></script>
<script type="text/javascript" src="_static/underscore.js"></script>
<script type="text/javascript" src="_static/doctools.js"></script>
<link rel="shortcut icon" href="_static/favicon.ico"/>
<link rel="author" title="About these documents" href="about.html" />
<link rel="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
<link rel="next" title="About Darkon" href="about.html" />
<meta property="og:site_name" content="darkon" />
<meta property="og:title" content="Welcome to Darkon" />
<meta property="og:description" content="Toolkit to Hack Your Deep Learning Models." />
<meta property="og:image" content="http://darkon.io/_static/logo-og.png" />
<meta property="og:url" content="http://darkon.io" />
<meta property="og:type" content="website" />
<meta name="twitter:card" content="summary_large_image" />
<link rel="stylesheet" href="_static/custom.css" type="text/css" />
<meta name="viewport" content="width=device-width, initial-scale=0.9, maximum-scale=0.9" />
</head>
<body>
<div class="document">
<div class="sphinxsidebar" role="navigation" aria-label="main navigation">
<div class="sphinxsidebarwrapper">
<p class="logo">
<a href="#">
<img class="logo" src="_static/logo.png" alt="Logo"/>
</a>
</p>
<p class="blurb">Toolkit to Hack Your Deep Learning Models.</p>
<p>
<iframe src="https://ghbtns.com/github-btn.html?user=darkonhub&repo=darkon&type=star&count=true&size=large&v=2"
allowtransparency="true" frameborder="0" scrolling="0" width="200px" height="35px"></iframe>
</p>
<h3>Navigation</h3>
<ul>
<li class="toctree-l1"><a class="reference internal" href="about.html">About Darkon</a></li>
<li class="toctree-l1"><a class="reference internal" href="api/index.html">API</a></li>
<li class="toctree-l1"><a class="reference internal" href="example.html">Examples</a></li>
<li class="toctree-l1"><a class="reference internal" href="version.html">Versions</a></li>
</ul>
<div class="relations">
<h3>Related Topics</h3>
<ul>
<li><a href="#">Documentation overview</a><ul>
<li>Next: <a href="about.html" title="next chapter">About Darkon</a></li>
</ul></li>
</ul>
</div>
<div id="searchbox" style="display: none" role="search">
<h3>Quick search</h3>
<form class="search" action="search.html" method="get">
<div><input type="text" name="q" /></div>
<div><input type="submit" value="Go" /></div>
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
<script type="text/javascript">$('#searchbox').show(0);</script>
</div>
</div>
<div class="documentwrapper">
<div class="bodywrapper">
<div class="body" role="main">
<div class="section" id="darkon-documentation">
<h1>Darkon Documentation<a class="headerlink" href="#darkon-documentation" title="Permalink to this headline">¶</a></h1>
<div class="section" id="about-darkon">
<h2>About Darkon<a class="headerlink" href="#about-darkon" title="Permalink to this headline">¶</a></h2>
<a class="reference external image-reference" href="https://travis-ci.org/darkonhub/darkon"><img alt="Build Status" src="https://travis-ci.org/darkonhub/darkon.svg?branch=master" /></a>
<a class="reference external image-reference" href="https://codecov.io/gh/darkonhub/darkon"><img alt="codecov" src="https://codecov.io/gh/darkonhub/darkon/branch/master/graph/badge.svg" /></a>
<a class="reference external image-reference" href="https://opensource.org/licenses/Apache-2.0"><img alt="License" src="https://img.shields.io/badge/License-Apache%202.0-blue.svg" /></a>
<a class="reference external image-reference" href="https://pypi.python.org/pypi/darkon"><img alt="PyPI" src="https://img.shields.io/pypi/v/darkon.svg?style=flat-square" /></a>
<a class="reference external image-reference" href="https://gitter.im/darkonhub/darkon?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge"><img alt="Gitter" src="https://badges.gitter.im/darkonhub/darkon.svg" /></a>
<a class="reference external image-reference" href="https://www.codacy.com/app/zironycho/darkon?utm_source=github.com&amp;utm_medium=referral&amp;utm_content=darkonhub/darkon&amp;utm_campaign=Badge_Grade"><img alt="Codacy Badge" src="https://api.codacy.com/project/badge/Grade/077f07f7a52b4d8186beee724ed19231" /></a>
<hr class="docutils" />
<p><strong>Darkon: Toolkit to Hack Your Deep Learning Models</strong></p>
<p><strong>Darkon</strong> is an open source toolkit to understand deep learning models better. Deep learning is often referred as a black-box that is difficult to understand.
But, accountability and controllability could be critical to commercialize deep learning models. People often think that high accuracy on prepared dataset
is enough to use the model for commercial products. However, well-performing models on prepared dataset often fail in real world usages and cause corner cases
to be fixed. Moreover, it is necessary to explain the result to trust the system in some applications such as medical diagnosis, financial decisions, etc. We hope<span class="raw-html-m2r"><br></span>
<strong>Darkon</strong> can help you to understand the trained models, which could be used to debug failures, interpret decisions, and so on.</p>
<p>Here, we provide functions to analyze deep learning model decisions easily applicable to any Tensorflow models (other models to be supported later).
Influence score can be useful to understand the model through training samples. The score can be used for filtering bad training samples that affects test performance negatively.
It is useful to prioritize potential mislabeled examples to be fixed, and debug distribution mismatch between train and test samples.
In this version, we have added Grad-CAM and Guided Grad-CAM, which are useful to understand decisions of CNN models.</p>
<p>We will gradually enable technologies to analyze deep learning models easily applicable to your existing projects.
More features will be released soon. Feedback and feature request are always welcome, which help us to manage priorities. Please keep your eyes on <strong>Darkon</strong>.</p>
<div class="section" id="demo">
<h3>Demo<a class="headerlink" href="#demo" title="Permalink to this headline">¶</a></h3>
<p><a class="reference external" href="https://darkon-demo.herokuapp.com">Demo Page</a></p>
</div>
<div class="section" id="dependencies">
<h3>Dependencies<a class="headerlink" href="#dependencies" title="Permalink to this headline">¶</a></h3>
<ul class="simple">
<li><a class="reference external" href="https://github.com/tensorflow/tensorflow">Tensorflow</a>>=1.3.0</li>
</ul>
</div>
<div class="section" id="installation">
<h3>Installation<a class="headerlink" href="#installation" title="Permalink to this headline">¶</a></h3>
<p>Install Darkon alone</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>pip install darkon
</pre></div>
</div>
<p>Install with TensorFlow CPU</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>pip install darkon<span class="o">[</span>tensorflow<span class="o">]</span>
</pre></div>
</div>
<p>Install with TensorFlow GPU</p>
<div class="highlight-bash"><div class="highlight"><pre><span></span>pip install darkon<span class="o">[</span>tensorflow-gpu<span class="o">]</span>
</pre></div>
</div>
</div>
<div class="section" id="communication">
<h3>Communication<a class="headerlink" href="#communication" title="Permalink to this headline">¶</a></h3>
<ul class="simple">
<li><a class="reference external" href="https://github.com/darkonhub/darkon/issues">Issues</a>: report issues, bugs, and request new features</li>
<li><a class="reference external" href="https://github.com/darkonhub/darkon/pulls">Pull request</a></li>
<li>Discuss: <a class="reference external" href="https://gitter.im/darkonhub/darkon?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge">Gitter</a></li>
<li>Email: <a class="reference external" href="mailto:contact%40darkon.io">contact<span>@</span>darkon<span>.</span>io</a></li>
</ul>
</div>
<div class="section" id="authors">
<h3>Authors<a class="headerlink" href="#authors" title="Permalink to this headline">¶</a></h3>
<p><a class="reference external" href="http://www.neosapience.com">Neosapience, Inc.</a></p>
</div>
<div class="section" id="license">
<h3>License<a class="headerlink" href="#license" title="Permalink to this headline">¶</a></h3>
<p>Apache License 2.0</p>
</div>
<div class="section" id="references">
<h3>References<a class="headerlink" href="#references" title="Permalink to this headline">¶</a></h3>
<p>[1] Cook, R. D. and Weisberg, S. “<a class="reference external" href="https://www.casact.org/pubs/proceed/proceed94/94123.pdf">Residuals and influence in regression</a>”, New York: Chapman and Hall, 1982</p>
<p>[2] Koh, P. W. and Liang, P. “<a class="reference external" href="https://arxiv.org/abs/1703.04730">Understanding Black-box Predictions via Influence Functions</a>” ICML2017</p>
<p>[3] Pearlmutter, B. A. “<a class="reference external" href="http://www.bcl.hamilton.ie/~barak/papers/nc-hessian.pdf">Fast exact multiplication by the hessian</a>” Neural Computation, 1994</p>
<p>[4] Agarwal, N., Bullins, B., and Hazan, E. “<a class="reference external" href="https://arxiv.org/abs/1602.03943">Second order stochastic optimization in linear time</a>” arXiv preprint arXiv:1602.03943</p>
<p>[5] Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, Dhruv Batra “<a class="reference external" href="https://arxiv.org/abs/1610.02391">Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization</a>” ICCV2017</p>
<div class="toctree-wrapper compound">
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<div class="clearer"></div>
</div>
<div class="footer">
©<a href='http://www.neosapience.com'>2017 Neosapience, Inc</a>.
|
Powered by <a href="http://sphinx-doc.org/">Sphinx 1.6.6</a>
& <a href="https://github.com/bitprophet/alabaster">Alabaster 0.7.10</a>
|
<a href="_sources/index.rst.txt"
rel="nofollow">Page source</a>
</div>
<a href="https://github.com/darkonhub/darkon" class="github">
<img style="position: absolute; top: 0; right: 0; border: 0;" src="https://s3.amazonaws.com/github/ribbons/forkme_right_darkblue_121621.png" alt="Fork me on GitHub" class="github"/>
</a>
<script type="text/javascript">
var _gaq = _gaq || [];
_gaq.push(['_setAccount', 'UA-109936098-1']);
_gaq.push(['_setDomainName', 'none']);
_gaq.push(['_setAllowLinker', true]);
_gaq.push(['_trackPageview']);
(function() {
var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true;
ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js';
var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s);
})();
</script>
</body>
</html>