-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
executable file
·375 lines (309 loc) · 19.1 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
<!DOCTYPE HTML>
<html lang="bn"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<!-- Hi, Jon Here. Please DELETE the two <script> tags below if you use this HTML, otherwise my analytics will track your page -->
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=UA-7580334-2"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'UA-7580334-2');
</script>
<title>Yasiru Ranasinghe</title>
<meta name="author" content="Jon Barron">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" type="text/css" href="stylesheet.css">
<link rel="icon" type="image/png" href="images2/profile/sloth.png">
</head>
<body>
<table style="width:100%;max-width:800px;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr style="padding:0px">
<td style="padding:0px">
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr style="padding:0px">
<td style="padding:2.5%;width:63%;vertical-align:middle">
<p style="text-align:center">
<name>Yasiru Ranasinghe</name>
</p>
<p>
I am a 3<sup>rd</sup> year Ph.D. candidate at <a href="https://engineering.jhu.edu/vpatel36/">Vision & Image Understanding Lab</a>, <a href="https://www.jhu.edu/">Johns Hopkins University</a> under the supervision of <a href="https://scholar.google.com/citations?user=AkEXTbIAAAAJ&hl=en">Dr. Vishal M. Patel</a>. My research focuses on computer vision and its application on crowd analysis, object localization, and representation learning.
</p>
<p>
I worked as a research assistant for <a href="http://covid.eng.pdn.ac.lk/research.php">AI4COVID</a> project under the supervision of <a href="https://scholar.google.com/citations?user=yqup6Q8AAAAJ&hl=en">Prof. Janaka Ekanayake</a> funded by <a href="https://covidsouth.ai/">IDRC</a>. During my undergraduate at <a href="https://web2.ee.pdn.ac.lk/">University of Peradeniya</a>, Sri Lanka, I worked on computer vision and image processing applications on spectral imagery and remote sensing under the supervision of <a href="https://scholar.google.com/citations?user=6_XOJbsAAAAJ&hl=en">Prof. Roshan Godaliyadda</a>, <a href="https://scholar.google.com/citations?user=f5h5ByUAAAAJ&hl=en">Prof. Vijitha Herath</a>, and <a href="https://scholar.google.com/citations?user=uJvb7zwAAAAJ&hl=en">Prof. Parakrama Ekanayake</a>.
</p>
<p style="text-align:center">
<a href="mailto:dranasi1@jhu.edu">Email</a>  / 
<a href="https://scholar.google.com/citations?hl=en&user=sG77m5UAAAAJ">Google Scholar</a>  / 
<a href="https://www.linkedin.com/in/yasiruranasinghe/">LinkedIn</a>
</p>
</td>
<td style="padding:2.5%;width:40%;max-width:40%">
<a href="images2/profile/prof_image.png"><img style="width:100%;max-width:100%" alt="profile photo" src="images2/profile/prof_image.png" class="hoverZoomLink"></a>
</td>
</tr>
</tbody></table>
<hr>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:10px;width:100%;vertical-align:middle">
<heading>Research</heading>
<p>
My research interests primarily revolve around computer vision applications, strongly focusing on using generative models for downstream tasks such as
crowd counting, crowd localization, and monocular 3D object detection.
I am particularly interested in exploring the potential of diffusion models to enhance the accuracy and efficiency of these tasks, especially when applied to monocular images.
My overarching goal is to contribute to the advancement of computer vision techniques that can handle complex scenarios involving crowds and vehicles, with an emphasis on noise reduction,
localization, and object detection.
</p>
<p>
</p>
</td>
</tr>
</tbody></table>
<hr>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:20px;width:100%;vertical-align:middle">
<heading>News</heading>
<p>
<ul>
<li><b>March, 2024:</b> One Paper accepted at <b>IEEE FG'24</b>.</li>
<li><b>February, 2024:</b> Two Papers accepted at <b>CVPR'24</b>.</li>
<li><b>May, 2023:</b> One Paper accepted by <b>JFST</b>.</li>
<li><b>September, 2022:</b> Joined as a PhD student at <b>Vision & Image Understanding Lab, Johns Hopkins University</b>.</li>
<li><b>January, 2022:</b> One Paper is Accepted by <b>IEEE Access</b>.</li>
<li><b>December, 2021:</b> One Paper is Accepted by <b>IEEE JSTARS</b>.</li>
<!-- <li><b>November, 2020:</b> One Paper is Accepted at <b>ICIIS</b>.</li> -->
<!-- <li><b>December, 2019:</b> One Paper is Accepted at <b>ICIIS</b>.</li> -->
</ul>
</div>
</p>
</td>
</tr>
</tbody></table>
<table width="100%" align="center" border="0" cellpadding="0"><tbody>
<hr>
<heading>Selected Publications</heading>
<table width="100%" align="center" border="0" cellspacing="0" cellpadding="20">
<tr onmouseout="ff_stop()" onmouseover="ff_start()">
<td style="padding:20px;width:25%;vertical-align:middle">
<div class="one">
<div class="two" id='ff_image'>
<img src='images2/ddc/main_figure.png' height = "120" width="160"></div>
<img src='images2/ddc/flow_chart.jpg' height = "100" width="160"></div>
</div>
<script type="text/javascript">
function ff_start() {
document.getElementById('ff_image').style.opacity = "1";
}
function ff_stop() {
document.getElementById('ff_image').style.opacity = "0";
}
ff_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<p><a href="https://arxiv.org/pdf/2303.12790.pdf">
<papertitle><em>CrowdDiff</em>: Multi-hypothesis Crowd Density Estimation using Diffusion Models</a></papertitle>
<p>Authors : <strong>Yasiru Ranasinghe</strong>, Nithin Gopalakrishnan Nair, Wele Gedara Chaminda Bandara, and Vishal M. Patel</p>
<em>CVPR</em>, 2024   <br>
<!--<a href="https://arxiv.org/abs/2304.04745">Paper</a> /
<a href="https://aimansnigdha.github.io/cimd/">Website</a> /
<a href="https://github.com/aimansnigdha/Ambiguous-Medical-Image-Segmentation-using-Diffusion-Models">Code</a>-->
<p>The paper proposes a novel approach to perform crowd counting with multi-hypothesis aggregation using denoising diffusion probabilistic models. The approach outperforms existing state-of-the-art methods for crowd counting.
</p>
</a></p>
</td>
</tr>
<tr onmouseout="ff1_stop()" onmouseover="ff1_start()">
<td style="padding:20px;width:25%;vertical-align:middle">
<div class="one">
<div class="two" id='ff1_image'>
<img src='images2/monodiff/abstract_figure.jpg' height = "160" width="160"></div>
<img src='images2/monodiff/pipeline.jpg' height = "100" width="160"></div>
</div>
<script type="text/javascript">
function ff1_start() {
document.getElementById('ff1_image').style.opacity = "1";
}
function ff1_stop() {
document.getElementById('ff1_image').style.opacity = "0";
}
ff1_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<p><a href="https://drive.google.com/file/d/148LqyCjVPsr6WtPfhTQ3U6ESGtvPW2Ye/view?usp=drive_link">
<papertitle><em>MonoDiff</em>: Monocular 3D Object Detection and Pose Estimation with Diffusion Models</a></papertitle>
<p>Authors : <strong>Yasiru Ranasinghe</strong>, Deepti Hegde, and Vishal M. Patel</p>
<em>CVPR</em>, 2024   <br>
<!--<a href="https://arxiv.org/abs/2304.04745">Paper</a> /
<a href="https://aimansnigdha.github.io/cimd/">Website</a> /
<a href="https://github.com/aimansnigdha/Ambiguous-Medical-Image-Segmentation-using-Diffusion-Models">Code</a>-->
<p>The paper proposes using diffusion models to perform monocular 3D object detection and pose estimation. <em>MonoDiff</em>
does not require additional modalities to generate intermediate representations to produce box parameters.
</p>
</a></p>
</td>
</tr>
<tr onmouseout="ff0_stop()" onmouseover="ff0_start()">
<td style="padding:20px;width:25%;vertical-align:middle">
<div class="one">
<div class="two" id='ff0_image'>
<img src='images2/crowdloc/image1.png' height = "120" width="160"></div>
<img src='images2/crowdloc/image2.png' height = "32" width="160"></div>
</div>
<script type="text/javascript">
function ff0_start() {
document.getElementById('ff0_image').style.opacity = "1";
}
function ff0_stop() {
document.getElementById('ff0_image').style.opacity = "0";
}
ff0_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:middle">
<p><a href="https://drive.google.com/file/d/1nhNYMJ9XFatDRVcedLiEF6Ue-VjzVGDe/view?usp=drive_link">
<papertitle>Crowd Detection via Point Localization with Diffusion Models</a></papertitle>
<p>Authors : <strong>Yasiru Ranasinghe</strong>, and Vishal M. Patel</p>
<em>IEEE FG</em>, 2024   <br>
<!--<a href="https://arxiv.org/abs/2304.04745">Paper</a> /
<a href="https://aimansnigdha.github.io/cimd/">Website</a> /
<a href="https://github.com/aimansnigdha/Ambiguous-Medical-Image-Segmentation-using-Diffusion-Models">Code</a>-->
<p>The paper proposes using diffusion models to localize crowd in images and count the number of people in a scene.
The proposed method generates the head locations as a generative task without a separate detector for point proposals.
</p>
</a></p>
</td>
</tr>
<tr onmouseout="ff2_stop()" onmouseover="ff2_start()">
<td style="padding:20px;width:25%;vertical-align:top">
<div class="one">
<div class="two" id='ff2_image'>
<img src='images2/access/results.png' height="90" width="160"></div>
<img src='images2/access/main_figure.png' height="80" width="160">
</div>
<script type="text/javascript">
function ff2_start() {
document.getElementById('ff2_image').style.opacity = "1";
}
function ff2_stop() {
document.getElementById('ff2_image').style.opacity = "0";
}
ff2_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:top">
<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9686744">
<papertitle>Transmittance Multispectral Imaging for Reheated Coconut Oil Differentiation</a></papertitle>
<p>Authors : <strong>DYL Ranasinghe </strong>, HK Weerasooriya, S Herath, MP Bandara Ekanayake, HMVR Herath, GMRI Godaliyadda, and Terrence Madhujith</p>
<em>IEEE Access</em>, 2022   <br>
<p>The article process a novel non invasive method for food quality analysis, specifically in terms of adulteration, using multispectral images and statistical signal processing and image processing. The proposed method yielded statistically significant results and with significant practical implications.</p>
<p></p>
</a></p>
</td>
</tr>
<tr onmouseout="ff5_stop()" onmouseover="ff5_start()">
<td style="padding:20px;width:25%;vertical-align:top">
<div class="one">
<div class="two" id='ff5_image'>
<img src='images2/jstars/algorithm.png' height="180" width="160"></div>
<img src='images2/jstars/main_figure.png' height="100" width="160">
</div>
<script type="text/javascript">
function ff5_start() {
document.getElementById('ff5_image').style.opacity = "1";
}
function ff5_stop() {
document.getElementById('ff5_image').style.opacity = "0";
}
ff5_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:top">
<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9609564">
<papertitle>Constrained Nonnegative Matrix Factorization for Blind Hyperspectral Unmixing Incorporating Endmember Independence</a></papertitle>
<p>Authors : EMMB Ekanayake, HMHK Weerasooriya, <strong>DYL Ranasinghe </strong>, S Herath, B Rathnayake, GMRI Godaliyadda, MPB Ekanayake, and HMVR Herath</p>
<em>IEEE Journal of Seleted Topics in Applied Earth Observation and Remote Sensing</em>, 2021   <br>
<p>
The paper proposes a novel algorithm to extract endmembers and abundances of hyperspectral remote sensing data. We introduced a regularizer for the nonnegative matrix factorization which improves independence of endmember statistics. The algorithm illustrated superior performance interms of endmber extraction from hyperspectral data.
</p>
<p></p>
</a></p>
</td>
</tr>
<tr onmouseout="ff3_stop()" onmouseover="ff3_start()">
<td style="padding:20px;width:25%;vertical-align:top">
<div class="one">
<div class="two" id='ff3_image'>
<img src='images2/iciis_2020/results.png' height="180" width="160"></div>
<img src='images2/iciis_2020/main_figure.png' height="100" width="160">
</div>
<script type="text/javascript">
function ff3_start() {
document.getElementById('ff3_image').style.opacity = "1";
}
function ff3_stop() {
document.getElementById('ff3_image').style.opacity = "0";
}
ff3_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:top">
<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9342727">
<papertitle>Convolutional Autoencoder for Blind Hyperspectrl Unmixing</a></papertitle>
<p>Authors : <strong>Yasiru Ranasinghe</strong>, Sanjaya Herath, Kavinga Weerasooriya, Mevan Ekanayake, Roshan Godaliyadda, Parakrama Ekanayake, and Vijitha Herath</p>
<em>IEEE International Conference on Industrial and Information Systems</em>, 2020   <br>
<p>
The paper process a convolutional autoencoder to realize the nonnegative matrix factorization using deep learning architectures. The proposed methods is applied to extract endmembers and abundances of hyperspectral remote sensing data. The proposed method produced state-of-the-art results on abundance estimation and competitive results in terms of endmember extraction.
</p>
<p></p>
</a></p>
</td>
</tr>
<tr onmouseout="ff4_stop()" onmouseover="ff4_start()">
<td style="padding:20px;width:25%;vertical-align:top">
<div class="one">
<div class="two" id='ff4_image'>
<img src='images2/iciis_2019/results.png' height="200" width="160"></div>
<img src='images2/iciis_2019/main_figure.png' height="160" width="160">
</div>
<script type="text/javascript">
function ff4_start() {
document.getElementById('ff4_image').style.opacity = "1";
}
function ff4_stop() {
document.getElementById('ff4_image').style.opacity = "0";
}
ff4_stop()
</script>
</td>
<td style="padding:20px;width:75%;vertical-align:top">
<a href="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9063280">
<papertitle>Hyperspectral Imaging Based Method to Identify Potential Limestone Deposis</a></papertitle>
<p>Authors : <strong>DYL Ranasinghe</strong>, HMS Lakmal, HMHK Weerasooriya, EMMB Ekanayake, GMRI Godaliyadda, HMVR Herath, and MPB Ekanayake</p>
<em>IEEE International Conference on Industrial and Information Systems</em>, 2019   <br>
<p>
The paper proposed an algorithm to determine the availability of surface limestone using hyperspectral satellite imagery of an area. We incorporate traditional image and signal processing, and statistical data analysis techniques in the algorithm. Further, we generated a self-supervisied representation for the hyperspectral signature of limestone in the absence of a groundtruth to improve classification accuracy and spatial continuity of the probability map.
</p>
<p></p>
</a></p>
</td>
</tr>
</tbody></table>
<hr>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:0px">
<br>
<p style="text-align:right;font-size:small;">
Template taken from <a href="https://github.com/jonbarron/jonbarron_website">here</a>. Last updated October 2023.
</p>
</td>
</tr>
</tbody></table>
</td>
</tr>
</table>
</body>
</html>