-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
406 lines (319 loc) · 18.8 KB
/
index.html
File metadata and controls
406 lines (319 loc) · 18.8 KB
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
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
<!DOCTYPE HTML>
<html lang="en">
<head>
<meta charset="UTF-8">
<title>Eric Zhao</title>
<meta name="author" content="Eric Zhao">
<meta name="viewport" content="width=device-width,initial-scale=1">
<link rel="stylesheet" href="stylesheet.css">
<link rel="icon" href="images/icons/rocket.svg">
</head>
<body>
<!-- 1) Tab buttons -->
<div class="tab">
<button class="tablinks" onclick="openTab(event,'Home')" id="defaultOpen">Home</button>
<button class="tablinks" onclick="openTab(event,'Projects')">Projects</button>
<button class="tablinks" onclick="openTab(event,'Research')">Research</button>
<button class="tablinks" onclick="openTab(event,'Hobbies')">Hobbies</button>
</div>
<!-- 2) Home content -->
<div id="Home" class="tabcontent">
<!-- your existing homepage HTML goes here -->
<!-- … the <table> with About Me / Research list … -->
<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:60%;vertical-align:middle">
<p style="text-align:center">
<name>Eric Zhao</name>
</p>
<p>
I am a current M.S. student in Artificial Intelligence Engineering at Carnegie Mellon University, and a <a href="https://www.cmu.edu/graduate/rales-fellows/">Rales Fellow</a> at CMU, a fully funded graduate fellowship for exceptional leaders and innovators in STEM fields.
</p>
<p>
I graduated from Carnegie Mellon University with my B.S. in Mechanical Engineering with an Additional Major in Chinese Studies in 2024, where I was also selected as a <a href="https://www.thegatesscholarship.org/scholarship">Gates Scholar (Cohort III)</a>, and a <a href="https://www.cmu.edu/student-success/programs/tartan-scholars.html">Tartan Scholar</a>.
</p>
<p>
I'm interested in Reinforcement Learning, specifically in the field of controls and robotics. I'm currently working with the <a href="https://safeai-lab.github.io/">Safe AI Lab</a> in integrating LLMs for context awareness and semantic understanding of environments as an extension of their recent research on Multi-Agent locomtion and manipulation, or <a href="https://collaborative-mapush.github.io/">MAPush</a>.
</p>
<p>
I'm also developing a learning-based control pipeline for transtibial assistive devices, leveraging Reinforcement Learning and Imitation Learning techniques to train humanoid locomotion models in MuJoCo. This work is being conducted under the guidance of Dr. Inseung Kang, director of the <a href="https://metamobility.cmu.edu/">MetaMobility Lab</a> at CMU.
</p>
<p>
In my undergraduate studies at CMU, I was an active research in biomedical and mechanical engineering as a part of the Computational Engineering and Robotics Lab (CERLAB), directed by Dr. Kenji Shimada. I've conducted studies to determine the relationship between the physical parameters and resulting mechanical properties of lattice structures, and worked with our lab's novel mesh generation software to develop anisotropic lattices for customized transtibial prosthetic liners.
</p>
<p>
I served as a grader for several undergraduate engineering courses, including Mechanical Design and Numerical Methods.
</p>
<p style="text-align:center">
<a href="mailto:ericzhao@andrew.cmu.edu">Email</a>  / 
<a href="data/Eric_Zhao_AI_ML_Resume.pdf">CV</a>  / 
<a href="https://www.linkedin.com/in/eric-zhao2/">Linkedin</a>  / 
<a href="https://github.com/EZ-Blue/">Github</a>
</p>
</td>
<td style="padding:2.5%;width:40%;max-width:40%">
<a href="data/ez.jpg"><img style="width:100%;max-width:100%" alt="profile photo" src="data/ez.jpg" class="hoverZoomLink"></a>
</td>
</tr>
</tbody>
</table>
<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>Highlighted Work</heading>
<p>
I'm interested in reinforcement learning, robotics, and product design. I like optimization and working on research and projects that have a tangible impact in industry and in society.
</p>
</td>
</tr>
</tbody>
</table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tbody>
<tr bgcolor=#eee>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="images/demo.gif" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="research/locomotion.html">
<papertitle>Imitation and Reinforcement Learning for Optimal Ankle Prosthesis Control</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/manuellancastre/"> Manel Lancastre</a>,
<a href="https://www.linkedin.com/in/jonathan-he-628493248/"> Jonathan He</a>,
<a href="https://www.linkedin.com/in/shawnkrishnan/"> Shawn Krishnan</a>
<br>
<em>in review</em> / <a href="https://ez-blue.github.io/ILRL_prosthesis_control/"> Website</a>
<br>
<!-- <a href="https://model-based-plugging.github.io">website</a> /
<a href="https://arxiv.org/abs/2312.09190">arXiv</a> -->
<p>
A comparative study of imitation learning and reinforcement learning techniques in training effective ankle prosthetic controllers for plausible human locomotion, simulated using MuJoCo.
</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="images/tinypointnext.png" alt="tinypointnext" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="projects/TinyPointNext.html">
<papertitle>TinyPointNeXt: Reducing Model Parameters for Efficient Dynamic Object Segmentation</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/rohxnsxngh/"> Rohan Singh</a>,
<a href="https://www.linkedin.com/in/dylan-k-leong/"> Dylan Leong</a>
<br>
<a href="https://github.com/dylan813/tinypointnext">GitHub</a> /
<a href="projects/data/TinyPointNext Report.pdf">Report</a>
<p>
An optimized variant of PointNeXt architecture achieving 93.86% accuracy for human detection in point clouds while reducing model parameters by 81% and inference time by 17%, enabling real-time deployment on resource-constrained devices.
</p>
</td>
</tr>
<tr bgcolor=#eee>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="images/mapush.png" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="">
<papertitle>Multi-Agent Reinforcement Learning with LLMs for Safe Path Planning</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/gravesreid/"> Reid Graves</a>,
<a href="https://www.linkedin.com/in/michael-chase-allen/"> Michael Chase Allen</a>,
<a href="https://www.linkedin.com/in/jonathanali2025/"> Jonathan Ali</a>,
<a href="https://www.linkedin.com/in/shawnkrishnan/"> Shawn Krishnan</a>
<br>
<em>in review</em>
<br>
<a href="https://collaborative-mapush.github.io/">website</a>
<p>
A framework for context-aware path planning in multi-agent loco-manipulation scenarios through integration of LLM reasoning to an existing hierarchical multi-agent planning method. Demonstrates improved semantic understanding of obstacles in the coordination of quadruped robots in navigation and simple pushing tasks.
</p>
</td>
</tr>
</tbody>
</table>
</td>
</tr>
</tbody>
</table>
</div>
<!-- 3) Projects content -->
<div id="Projects" class="tabcontent">
<h2>Projects</h2>
<p>My projects span AI and Machine Learning, product design, and biomedical engineering.</p>
<!-- e.g. a grid of cards linking to GitHub demos -->
<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:25%;vertical-align:middle">
<img src="images/tinypointnext.png" alt="tinypointnext" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="projects/TinyPointNext.html">
<papertitle>TinyPointNeXt: Reducing Model Parameters for Efficient Dynamic Object Segmentation</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/rohxnsxngh/"> Rohan Singh</a>,
<a href="https://www.linkedin.com/in/dylan-k-leong/"> Dylan Leong</a>
<br>
<a href="https://github.com/dylan813/tinypointnext">GitHub</a> /
<a href="projects/data/TinyPointNext Report.pdf">Report</a>
<p>
An optimized variant of PointNeXt architecture achieving 93.86% accuracy for human detection in point clouds while reducing model parameters by 81% and inference time by 17%, enabling real-time deployment on resource-constrained devices.
</p>
</td>
</tr>
<tr bgcolor=#eee>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="projects/images/fifa_dataset.png" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="projects/fifa.html">
<papertitle>FIFA Players Analysis: Predictive Modeling of Player Ratings Using Machine Learning</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.cmu.edu/graduate/rales-fellows/bios/nicole-villavicencio-garduno"> Nicole Villavicencio Garduno</a>
<br>
<a href="https://github.com/Systems-and-Toolchains-Fall-2024/course-project-option-i-EZ-Blue.git">GitHub</a>
<p>
Comparing machine learning models to predict player ratings across multiple years of FIFA player data, testing implementations of linear regression, random forest, MLP and CNN models. Achieved over 90% accuracy in player overall value prediction with all models.
</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="projects/images/deliverulator_1.png" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="projects/deliverulator.html">
<papertitle>Delivery System Emulator (Deliverulator)</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/caseywalker/"> Casey Walker</a>,
<a href="https://www.linkedin.com/in/atharva-mhaskar/"> Atharva Mhaskar</a>,
<a href="https://www.linkedin.com/in/mya-chappell-4149b61bb/"> Mya Chappell</a>,
<a href="https://www.linkedin.com/in/rubin-chen/"> Rubin Chen</a>
<br>
<a href="https://github.com/kcmw3e/deliverulator/tree/trunk">Report</a>
<p>
This project is a simulated delivery system/game in which a player selects from a set of available delivery robots and assigns them to randomly-generated orders. Each robot follows a planned path in order to accomplish its delivery around locations on CMU's campus.
</p>
</td>
</tr>
<tr bgcolor=#eee>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="projects/images/ezsit_cover.png" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="projects/EZSit.html">
<papertitle>EZ Sit: An Ergonomic Standing Wheelchair with Sitting and Standing Assistance</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/christopher-oh-506a9218a/"> Chris Oh</a>,
<a href="https://www.linkedin.com/in/jingbo23zhang/"> Jingbo Zhang</a>,
<a href="https://www.yhwu33.com/"> Yunhuan Wu</a>
<br>
<a href="projects/data/EZSit Final Report.pdf">Report</a>
<p>
Senior capstone project for Engineering Product Design. Created an ergonomic standing wheelchair with sitting and standing assistance for individuals with mobility issues.
</p>
</td>
</tr>
</tbody>
</table>
</div>
<!-- 4) Research content -->
<div id="Research" class="tabcontent">
<h2>Research</h2>
<!-- you can copy your existing Research <table> into here -->
<p>
I'm interested in reinforcement learning, robotics, and product design. I like optimization and working on research and projects that have a tangible impact in industry and in society.
</p>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tbody>
<tr bgcolor=#eee>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="images/demo.gif" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="research/locomotion.html">
<papertitle>Imitation and Reinforcement Learning for Optimal Ankle Prosthesis Control</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/manuellancastre/"> Manel Lancastre</a>,
<a href="https://www.linkedin.com/in/jonathan-he-628493248/"> Jonathan He</a>,
<a href="https://www.linkedin.com/in/shawnkrishnan/"> Shawn Krishnan</a>
<br>
<em>in review</em> / <a href="https://ez-blue.github.io/ILRL_prosthesis_control/"> Website</a>
<br>
<!-- <a href="https://model-based-plugging.github.io">website</a> /
<a href="https://arxiv.org/abs/2312.09190">arXiv</a> -->
<p>
A comparative study of imitation learning and reinforcement learning techniques in training effective ankle prosthetic controllers for plausible human locomotion, simulated using MuJoCo.
</p>
</td>
</tr>
<tr>
<td style="padding:20px;width:25%;vertical-align:middle">
<img src="images/mapush.png" alt="b3do" width="160" style="border-style: none">
</td>
<td width="75%" valign="middle">
<a href="">
<papertitle>Multi-Agent Reinforcement Learning with LLMs for Safe Path Planning</papertitle>
</a>
<br>
<strong>Eric Zhao</strong>,
<a href="https://www.linkedin.com/in/gravesreid/"> Reid Graves</a>,
<a href="https://www.linkedin.com/in/michael-chase-allen/"> Michael Chase Allen</a>,
<a href="https://www.linkedin.com/in/jonathanali2025/"> Jonathan Ali</a>,
<a href="https://www.linkedin.com/in/shawnkrishnan/"> Shawn Krishnan</a>
<br>
<em>in review</em>
<br>
<a href="https://collaborative-mapush.github.io/">website</a>
<p>
A framework for context-aware path planning in multi-agent loco-manipulation scenarios through integration of LLM reasoning to an existing hierarchical multi-agent planning method. Demonstrates improved semantic understanding of obstacles in the coordination of quadruped robots in navigation and simple pushing tasks.
</p>
</td>
</tr>
</tbody>
</table>
</div>
<!-- 5) Hobbies & Interests -->
<div id="Hobbies" class="tabcontent">
<h2>Hobbies & Interests</h2>
<p>Basketball, music, cooking, baking.</p>
</div>
<!-- 6) Tab‐switching script -->
<script>
function openTab(evt, tabName) {
// hide all .tabcontent
document.querySelectorAll('.tabcontent').forEach(tc => tc.style.display = 'none');
// remove “active” class from all buttons
document.querySelectorAll('.tablinks').forEach(btn => btn.classList.remove('active'));
// show the one we clicked
document.getElementById(tabName).style.display = 'block';
evt.currentTarget.classList.add('active');
}
// open Home by default
document.getElementById('defaultOpen').click();
</script>
</body>
</html>