-
Notifications
You must be signed in to change notification settings - Fork 2
/
index.html
391 lines (340 loc) · 22.2 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
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
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
<!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">
<link href="https://fonts.cdnfonts.com/css/chalkduster" rel="stylesheet">
<style>
@import url('https://fonts.cdnfonts.com/css/chalkduster');
</style>
<!-- <script src="./sm/assets/teaser-data.js"></script> -->
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>Image Retrieval</title>
<link href="style.css" rel="stylesheet" type="text/css">
<meta name="description"
content="Project page for 'Composite Sketch+Text Queries for Retrieving Objects with Elusive Names and Complex Interactions.'">
<!-- <link rel="icon" href="./pics/wis_logo.jpg"> -->
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css"
integrity="sha384-Gn5384xqQ1aoWXA+058RXPxPg6fy4IWvTNh0E263XmFcJlSAwiGgFAW/dAiS6JXm" crossorigin="anonymous">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>Image Retrieval</title>
<link href="styles/style.css" rel="stylesheet" type="text/css">
<meta name="description"
content="Project page for 'Composite Sketch+Text Queries for Retrieving Objects with Elusive Names and Complex Interactions.'">
<meta charset="utf-8">
<meta name="description" content="Composite Sketch+Text Queries for Retrieving Objects with Elusive Names and Complex Interactions">
<meta name="keywords" content="sketch+text-based image retrieval, cross-modal retrieval, image retrieval, SBIR, CSTBIR;">
<meta name="author" content="Prajwal Gatti">
<title>Image Retrieval</title>
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:title" content="[AAAI 2024] Composite Sketch+Text Queries for Retrieving Objects with Elusive Names and Complex Interactions" />
<meta name="twitter:description" content="[AAAI 2024] Composite Sketch+Text Queries for Retrieving Objects with Elusive Names and Complex Interactions" />
<meta name="twitter:image:alt" content="CSTBIR (AAAI 2024)" />
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/font-awesome/4.7.0/css/font-awesome.min.css">
</head>
<body>
<p class="title">Image Retrieval</p>
<div style="text-align: center; font-size: 15pt; margin-bottom: 15px">
<span><a href="https://drive.google.com/file/d/18GM2JR2IFbwIbOrfJeCXIvVkdYOiPdBt/view">CSL2050 Course Project 2024</a></span>
</div>
<p class="author">
<span class="author"><a href="https://github.com/Jay18Mehta"> Jay Mehta</a></span>
<span class="author"><a href="https://github.com/akshatb22me007">Akshat Jain</a></span>
<span class="author"><a href="https://github.com/harshivshah2504">Harshiv Shah</a></span>
<span class="author"><a href="https://github.com/gjyotin305">Jyotin Goel</a></span>
<span class="author"><a href="https://github.com/RHYTHM2405">Rhythm Baghel</a></span>
</p>
<table border="0" align="center" class="affiliations" width="700px">
<tbody align="center">
<tc>
<td style="text-align: center;">IIT Jodhpur</td>
</tc>
</tbody>
</table>
<table width="999" border="0" align="center" class="menu" style="margin-bottom: 8px;">
<tbody>
<tr>
<td align="center">| <a href="https://drive.google.com/file/d/1D6Xfj3wfW33zipDI71W57nRNVdMl-7Ce/view">Report link</a> | <a href="https://github.com/gjyotin305/CSL2050_CourseProject">Github Link</a> | <a href="https://www.cs.toronto.edu/~kriz/cifar.html">Cifar-10 Dataset</a> | <a href="https://youtu.be/6uMFCbYbJjo?si=0IdReEe1-jd551PY">Youtube link</a> </td>
</tr>
</tbody>
</table>
<div class="container">
<table width="1000" border="0" align="center">
<tbody>
<tr>
<div id="imageCarousel" class="carousel slide teaser-carousel" data-ride="carousel" data-interval="5000">
<div class="carousel-inner">
<div class="carousel-item active">
<div>
<img class="teaser-img" src="images/harshiv.jpg" height="390" />
</div>
</div>
<div class="carousel-item">
<div>
<img class="teaser-img" src="images/truck.jpg" height="390" />
</div>
</div>
<div class="carousel-item">
<div>
<img class="teaser-img" src="images/plane.jpg" height="390" />
</div>
</div>
<div class="carousel-item">
<div>
<img class="teaser-img" src="images/horse.jpg" height="390" />
</div>
</div>
</div>
<a class="carousel-control-prev" href="#imageCarousel" role="button" data-slide="prev">
<div class="slider-navigation-previous">
<svg viewBox="0 0 50 80" xml:space="preserve">
<polyline fill="white" stroke-width=".5em" stroke-linecap="round"
stroke-linejoin="round" points="45.63,75.8 0.375,38.087 45.63,0.375 ">
</polyline>
</svg>
</div>
<span class="sr-only">Previous</span>
</a>
<a class="carousel-control-next" href="#imageCarousel" role="button" data-slide="next">
<div class="slider-navigation-next">
<svg viewBox="0 0 50 80" xml:space="preserve">
<polyline fill="white" stroke-width=".5em" stroke-linecap="round"
stroke-linejoin="round" points="0.375,0.375 45.63,38.087 0.375,75.8 ">
</polyline>
</svg>
</div>
<span class="sr-only">Next</span>
</a>
</div>
</tr>
<!-- <tr> <br /> </tr> -->
<tr align="center"></tr>
</tbody>
</table>
<div style="text-align: justify;text-justify: inter-word;">
<br>
<br>
<p><span class="section"><b>Abstract</b></span> </p>
<p>This project focuses on image retrieval, aiming to retrieve relevant images given an image query. Leveraging both Histogram of Oriented Gradients (HoG) and Convolutional Neural Network (CNN) features extracted through provided implementations, various methodologies are explored, including classification and clustering and mean based techniques. The CIFAR-10 dataset serves as the foundation for this study, offering a diverse collection of images for training and evaluation purposes. Through systematic experimentation and analysis, this project seeks to enhance image retrieval systems, contributing to advancements in image recognition and retrieval technologies..</p>
</div>
<!-- <div style="text-align: justify;text-justify: inter-word;">
<p><b>Keywords:</b> sketch+text-based image retrieval, cross-modal retrieval, image retrieval, SBIR, CSTBIR</p>
</div> -->
<p class="section"> </p>
<p class="section"><b>The Image Retrieval Problem</b></p>
<div style="text-align: center">
<img src="./images/image2image.png" alt="" width="700px" style="margin: auto" />
<br><br><p>In today's digital age, the vast amount of image data available on the internet poses a significant challenge in efficiently retrieving relevant images based on user queries. To address this challenge, the task at hand is to develop an Image Retrieval System capable of retrieving relevant images given an image query.</p>
</div>
<p class="section"> </p>
<!-- <p class="section"><b>Sketches in CSTBIR</b></p>
<div style="text-align: center">
<img src="./resources/supp_sketch_examples-min.png" alt="" width="700px" style="margin: auto" /><br>
<br><br><p>Examples of sketches.</p>
</div>
<p class="section"> </p> -->
<p class="section"><b>Youtube Video</b></p>
<centre>
<iframe width="560" height="315" src="https://www.youtube.com/embed/6uMFCbYbJjo?si=uTvRotgH5AHOsw1V" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>
</centre>
<p class="section"> </p>
<div class="papers-list">
<p class="section" id="related-paper"><b>Our Approaches</b></p>
<ul>
<li>
<h5>KNN:</h5> Utilizes similarity of feature vectors for classification.
Without additional feature extraction techniques, KNN operates directly on raw pixel values or basic features.
</li><br>
<li>
<h5>PCA+KNN:</h5>Principal Component Analysis (PCA) reduces dimensionality to capture significant variations.By reducing the dimensionality of the feature space, PCA aims to capture the most significant variations in the data.
</li><br>
<li>
<h5>Quicknet + KNN:</h5>Used Quicknet for feature extraction of training images and testing images.Next used KNN on extracted features, which gives K nearest images from training data as output.This KNN modal has 72 percent accuracy on testing accuracy. Also,wherever the images are retrieved of the wrong class, these wrong class
images are internally similar to the original class. For example, If the
input image is of a cat, the output images will contain mostly cats
with some other animals like dogs or horses, but any vehicle won't be
output
</li><br>
<li>
<h5>ANN+KNN:</h5> First trained ANN with ReLU activation function and SGD optimizer to classify input images. Extracted images from training data which has the same class as the predicted class. Applied KNN to find Similar images. Ann has 52 percent accuracy, so retrieved images will have 52 percent accuracy as well.
</li><br>
<li>
<h5>HOG+KNN:</h5> HOG captures object shape and appearance.Combined with KNN for similarity matching
</li><br>
<li>
<h5>PCA+HOG+KNN:</h5> Combines PCA dimensionality reduction with HOG's object shape capturing.PCA reduces the dimensionality of the feature space, while HOG descriptors capture object shapes and appearances.
</li><br>
<li>
<h5>Riyalnet + Euclidean distance:</h5>Used the pretrained Resnet50, along with 3 additional 3 hidden linear
layers to reduce the output layers size from 1000 to 10.
Trained the above model on CIFAR-10 dataset to fine the model to
give required output.
Used the fine-tuned model to predict the class of the input image and
then from the predicted class, used euclidian distance to retrieve the
images closest to the query image .
</li><br>
<li>
<h5>Quicknet + Centroid:</h5> Used Quicknet architecture to extract the features of CIFAR-10 dataset.
Now we calculated the mean embeddings of each class by calculating
the mean using each sample of the respective class.(The mean embed-
dings are referred as centroid)
After calculating 10 centroids (one for each class), we extracted the
features of the query image and compared the embeddings of the query
image with each centroid. The query image belongs to the class whose
centroid is nearest to it.
This method has an accuracy of 80 percent. This method is extremely
fast as it requires only 10 comparisons after the features have been
extracted..
</li><br>
</ul>
</div>
<p class="section"> </p>
<div class="team">
<p class="section" id="team"><b>Team</b></p>
<div class="row justify-content-center">
<div class="col-xs-12 col-sm-6 col-md-6 col-lg-2">
<div class="mainflip">
<div class="card">
<div class="card-body text-center card-no-padding">
<p><img src="images/jay.jpeg" alt="card image" height="130" width="100"></p>
<h5 class="card-title">Jay <br> Mehta</h5>
<ul class="list-inline">
<li class="list-inline-item">
<a class="social-icon text-xs-center" target="_blank" href="https://www.linkedin.com/in/jay-mehta-86a745258/">
<i class="fa fa-linkedin"></i>
</a>
</li>
<li class="list-inline-item">
<a class="social-icon text-xs-center" target="_blank" href="https://jay18mehta.github.io/Portfolio/">
<i class="fa fa-user"></i>
</a>
</li>
</ul>
</div>
</div>
</div>
</div>
<div class="col-xs-12 col-sm-6 col-md-6 col-lg-2">
<div class="mainflip">
<div class="card">
<div class="card-body text-center card-no-padding">
<p><img src="images/akshat.jpeg" alt="card image" height="130" width="100"></p>
<h5 class="card-title">Akshat <br> Jain</h5>
<ul class="list-inline">
<li class="list-inline-item">
<li class="list-inline-item">
<a class="social-icon text-xs-center" target="_blank" href="https://www.linkedin.com/in/akshat-jain-2bba94280/">
<i class="fa fa-linkedin"></i>
</a>
</li>
</a>
</li>
</ul>
</div>
</div>
</div>
</div>
<div class="col-xs-12 col-sm-6 col-md-6 col-lg-2">
<div class="mainflip">
<div class="card">
<div class="card-body text-center card-no-padding">
<p><img src="images/dog.jpg" alt="card image" height="130" width="100"></p>
<h5 class="card-title">Harshiv Shah</h5>
<ul class="list-inline">
<li class="list-inline-item">
<a class="social-icon text-xs-center" target="_blank" href="https://www.linkedin.com/in/harshivshah27?utm_source=share&utm_campaign=share_via&utm_content=profile&utm_medium=android_app">
<i class="fa fa-linkedin"></i>
</a>
</li>
</ul>
</div>
</div>
</div>
</div>
<div class="col-xs-12 col-sm-6 col-md-6 col-lg-2">
<div class="mainflip">
<div class="card">
<div class="card-body text-center card-no-padding">
<p><img src="images/rhythm.jpeg" alt="card image" height="130" width="100"></p>
<h5 class="card-title">Rhythm <br> Baghel</h5>
<ul class="list-inline">
<li class="list-inline-item">
<a class="social-icon text-xs-center" target="_blank" href="https://www.linkedin.com/in/rhythm-baghel-80675a25a/">
<i class="fa fa-linkedin"></i>
</a>
</li>
</ul>
</div>
</div>
</div>
</div>
<div class="col-xs-12 col-sm-6 col-md-6 col-lg-2">
<div class="mainflip">
<div class="card">
<div class="card-body text-center card-no-padding">
<p><img src="images/jyotin.jpeg" alt="card image" height="130" width="100"></p>
<h5 class="card-title">Jyotin <br> Goel</h5>
<ul class="list-inline">
<li class="list-inline-item">
<a class="social-icon text-xs-center" target="_blank" href="https://www.linkedin.com/in/jyotin-goel-16924b263/">
<i class="fa fa-linkedin"></i>
</a>
</li>
</ul>
</div>
</div>
</div>
</div>
</div>
<p class="section"> </p>
<p class="section"> </p>
<div class="papers-list">
<p class="section" id="related-paper"><b>Citations</b></p>
<ol>
<li>ResNet50 Paper - <a href="https://arxiv.org/abs/1512.03385">https://arxiv.org/abs/1512.03385</a></li>
<li>Dr Anand Mishra's website - <a href="https://vl2g.github.io/projects/cstbir/">https://vl2g.github.io/projects/cstbir/</a></li>
<li>HOG - <a href="https://ieeexplore.ieee.org/document/1467360">https://ieeexplore.ieee.org/document/1467360</a></li>
<li>KNN - <a href="https://www.researchgate.net/publication/2948052_KNN_Model-Based_Approach_in_Classification">https://www.researchgate.net/publication/2948052_KNN_Model-Based_Approach_in_Classification</a></li>
<li>CNN - <a href="https://arxiv.org/abs/1511.08458">https://arxiv.org/abs/1511.08458</a></li>
<li>PCA - <a href="https://www.researchgate.net/publication/316652806_Principal_Component_Analysis">https://www.researchgate.net/publication/316652806_Principal_Component_Analysis</a></li>
</ol>
</div>
<p class="section"> </p>
<p class="section"> </p>
<div class="ack">
<p class="section" id="ack"><b>Acknowledgment</b></p>
We are very grateful to Dr Anand Mishra for providing us with this opportunity to work on this project. This project helped us to explore many different techniques for image retrieval which helped us strengthen our basics and learn some of the advanced concepts. We were able to get a hands on experience while working on this project which was very beneficial for all of us.
</div>
<p class="section"></p>
<p class="section"> </p>
<div class="contact">
<p class="section" id="related-paper"><b>Contact</b></p>
For questions, please contact <a href="https://github.com/gjyotin305" target="_blank">Jyotin Goel</a> or raise an issue on <a class="publink" href="https://github.com/gjyotin305/CSL2050_CourseProject" target="_blank" style="text-decoration: none">GitHub</a>.
</div>
<p class="section"></p>
<!-- Footer -->
<footer>
<p class="m-0 text-center">Copyright © This is an open source project.</p>
</footer>
</body>
<script>
function prompt_on(prompt_element) {
prompt_element.classList.add("caption-active");
}
function prompt_off(prompt_element) {
prompt_element.classList.remove("caption-active");
}
function toggle_prompt(active_prompt_id, inactive_prompt_ids, result_id) {
let active_prompt = document.getElementById(active_prompt_id);
prompt_on(active_prompt);
for (let i = 0; i < inactive_prompt_ids.length; i++) {
let inactive_prompt = document.getElementById(inactive_prompt_ids[i]);
prompt_off(inactive_prompt);
}
let result = document.getElementById(result_id);
result.src = file_paths[active_prompt_id];
}
</script>
<script src="https://code.jquery.com/jquery-3.2.1.slim.min.js" integrity="sha384-KJ3o2DKtIkvYIK3UENzmM7KCkRr/rE9/Qpg6aAZGJwFDMVNA/GpGFF93hXpG5KkN" crossorigin="anonymous"></script>
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/umd/popper.min.js" integrity="sha384-ApNbgh9B+Y1QKtv3Rn7W3mgPxhU9K/ScQsAP7hUibX39j7fakFPskvXusvfa0b4Q" crossorigin="anonymous"></script>
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/js/bootstrap.min.js" integrity="sha384-JZR6Spejh4U02d8jOt6vLEHfe/JQGiRRSQQxSfFWpi1MquVdAyjUar5+76PVCmYl" crossorigin="anonymous"></script>
</html>