Repository navigation
Expand file tree
/
Copy pathindex.html
More file actions
597 lines (533 loc) · 27.9 KB
/
Copy pathindex.html
File metadata and controls
597 lines (533 loc) · 27.9 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
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Haocheng Li - Homepage</title>
<link href="css/bootstrap.css" rel='stylesheet' type='text/css' />
<link rel="shortcut icon" type="image/x-icon" href="images/icon.png">
<link href="https://fonts.googleapis.com/css?family=Open+Sans" rel="stylesheet">
<!-- Bulma Version 0.7.5-->
<link rel="stylesheet" href="https://unpkg.com/bulma@0.7.5/css/bulma.min.css" />
<link href='css/style.css?_t=20200916' rel='stylesheet' type='text/css'>
<script defer src="font-awesome/js/brands.min.js"></script>
<script defer src="font-awesome/js/regular.min.js"></script>
<script defer src="font-awesome/js/fontawesome.min.js"></script>
<style>
.red {
color: rgb(238, 76, 44);
font-style: normal;
}
.blue {
color: rgb(238, 76, 44);
font-style: normal;
}
#intro {
margin-top: 0em !important;
}
.content h3 {
margin-bottom: 1em!important;
margin-top: 2em!important;
}
.content figure {
width: 90%;
display: flex;
align-items: center;
overflow: hidden;
margin-left: auto!important;
margin-right: auto!important;
}
.columns:not(:last-child) {
margin-bottom: 1.75rem!important;
}
#sidebar {
width: 75%;
}
@media screen and (min-width: 769px), print {
.column.is-2_5, .column.is-2-tablet {
flex: none;
width: 20%;
}
}
</style>
<script src="https://code.jquery.com/jquery-3.1.1.min.js" crossorigin="anonymous"></script>
</head>
<body>
<section class="section">
<div class="container">
<div class="columns">
<div class="column is-2_5">
<div class="sticky">
<figure class="image" style="width: 11.6rem; margin-top: 6px;">
<img src="images/misc/profile.jpg">
</figure>
<div class="content">
<h2 style="margin-top: 1em">Haocheng Li</h2>
<p>
<b><a href="https://www.mccormick.northwestern.edu/robotics/" target="_blank">MSR</a> Student</b><br/>
@ Northwestern University<br/>
Evanston, IL<br/>
haochengli2018@gmail.com
</p>
</div>
<!-- social networks -->
<div class="social">
<a href="mailto:haochengli2018@gmail.com" target="_blank">
<span class="fa-regular fa-envelope fa-2x" style="display:inline; text-decoration: none; color: #4a4a4a; "></span>
</a>
<a href="https://github.com/AdamLi111/" target="_blank">
<span class="fab fa-github fa-2x" style="display:inline; text-decoration: none; color: #4a4a4a; margin-left: 15px"></span>
</a>
<a href="https://www.linkedin.com/in/haocheng-li-aug27/" target="_blank">
<span class="fab fa-linkedin fa-2x" style="display:inline; text-decoration: none; color: #4a4a4a; margin-left: 15px"></span>
</a>
</div>
<div id="sidebar" class="menu sticky is-hidden-mobile">
<p class="menu-label"><b>Quick Links</b></p>
<ul class="menu-list">
<li><a href="#Intro">About Me</a></li>
<li><a href="#Professional Experience">Professional Experience</a></li>
<li><a href="#Research">Research</a></li>
<li><a href="#Academic Projects">Academic Projects</a></li>
<!-- <li><a href="#Experiences">Experiences</a></li> -->
<li><a href="#Professional Projects">Professional Projects</a></li>
<!-- <li><a href="#Service">Service</a></li> -->
<li><a href="#Misc">Misc.</a></li>
</ul>
</div>
</div>
</div>
<div class="column right-panel">
<div class="content">
<!--About Me-->
<h2 id="Intro">About Me</h2>
<p>
I'm currently a Master of Science in Robotics (MSR) student at Northwestern University, expecting to graduate in Fall/Winter 2027. I received my B.S. in Computer Science from Northeastern University's Khoury College of Computer Sciences. With strong interest and passion about Embodied AI, I aim to build <strong>Embodied Agents</strong>🤖 that interact with humans👫 in real-world🌏 settings.
<br>
<p><strong>Coursework: </strong> Embedded Systems <strong>|</strong> Microcontroller System Design <strong>|</strong> Machine Dynamics <strong>|</strong> Robotic Manipulation <strong>|</strong> Algorithms and Data <strong>|</strong> Computer Systems <strong>|</strong> Object-Oriented Design <strong>|</strong> Artificial Intelligence <strong>|</strong> Machine Learning <strong>|</strong> Software Engineering <strong>|</strong> Reinforcement Learning</p>
<br>
<!-- <b>If you are interested in purchasing ImagePipe, please contact me by my <a href="mailto:haochengli2018@gmail.com" target="_blank">email</a></b> -->
</p>
<h2 id="Professional Projects">Professional Projects</h2>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/publications/imagepipe.png">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b>ImagePipe — AI-powered tourist photo processing system</b><br>
</p>
<p>
An end-to-end computer vision pipeline that automates the curation, enhancement, and organization of tourist photos — combining YOLOv8 classification, InsightFace recognition, MediaPipe eye-state detection, and a custom U-Net generative eye-opening model.
</p>
<p>
<a href="imagepipe.html" class="button is-link is-outlined is-small"><strong>View full demo →</strong></a>
</p>
</div>
</div>
</article>
<br>
<!--Research-->
<h2 id="Research">Research</h2>
<p>
I'm currently exploring my research interests in areas such as <strong>Robotic Manipulation, 3D Vision, and Human-Robot Interaction</strong>.
<br>
</p>
<ul>
<li><strong>Slow the Dialogue, Not Just the Robot: Positive Friction for Reliable Grounding and Safe, Embodied Vision-Language Action.</strong>:<br>
<i><b>Haocheng Li</b>, Asteria Kaeberlein, Mert Inan, Malihe Alikhani</i><br>
<i>[Target venue: <b>TACL (Transactions of the Association for Computational Linguistics)</b>] -- In preparation for submission</i><br>
<i>[Oct 2025 -- Present]</i><br>
<!-- <a href="https://arxiv.org/abs/2509.19853" target="_blank">[Paper Arxiv link]</a> -->
</p>
<article class="columns">
<div class="column is-7">
<figure class="image">
<img src="images/publications/misty.png">
<br>
</div>
<div class="column">
<div class="content">
<ul>
<li>Built a <strong>2D coordinate-based world model</strong> with collision detection, hazard zones, and dynamic FOV simulation for reproducible evaluation of embodied dialogue policies across 46 structured task scenarios.</li>
<li>Engineered a <strong>hierarchical ambiguity detection protocol</strong> that prioritizes safety hazards over navigation and referential errors, reducing collision-related task failures through ordered friction triggers.</li>
<li>Implemented a <strong>vague-to-specific synthetic evaluation pipeline</strong> using GPT-4o-mini as a simulated user agent, enabling controlled large-scale benchmarking that isolated contributions of conversation history (+7%), prompt design (+15.5%), and friction (+14.5%).</li>
</ul>
<b></b>
<!-- <a href="https://arxiv.org/abs/" target="_blank">[Coming Soon]</a> -->
<!-- <p>This is an ongoing survey paper with awesome list repo.</p> -->
</p>
</div>
</div>
</article>
</ul>
</p>
<ul>
<li><strong>SAGE: State-Aware Guided End-to-End Policy for Multi-Stage Sequential Tasks via Hidden Markov Decision Process</strong>:<br>
<i>Binxu Wu, Tengfei Zhang, Chen Yang, Jiahao Wen, Yixuan Huang, <b>Haocheng Li</b>, Jingtian Ma, Zhen Chen, Jingyuan Wang</i><br>
<i>[Submitted at: <b>RA-L (IEEE Robotics and Automation Letters)</b>, Oct 2026] -- Under Review <br>
[Dec 2024 -- Sep 2026]</i><br>
<a href="https://arxiv.org/abs/2509.19853" target="_blank">[Paper Arxiv link]</a>
</p>
<article class="columns">
<div class="column is-5">
<figure class="image">
<img src="images/publications/SAGE.png">
</figure>
<figure class="image">
<img src="images/publications/SAGE.gif">
</figure>
<br>
</div>
<div class="column">
<div class="content">
<ul>
<li>Reproduced UMI successfully after extensive code review and debugging.</li>
<li>Conducted comprehensive research on SLAM (ORB SLAM3), leading to a significant improvement in our
dataset's SLAM success rate <strong>from 43.1% to 97.9%</strong></li>
<li>Refined the gripper’s 3D design in <strong>SolidWorks</strong>, significantly improving maximum grasping force and stability</li>
<li>Led literature reviews and identified model improvements, actively driving research team's progress. Contributed significantly
to the <strong>Related Works</strong> section of our paper
</li>
<li>Reproduced Diffusion Policy's <strong>behavior cloning</strong> module and evaluated its performance and suitability for our model. </li>
<li>Collaborated on the development and debugging of the model's <strong>state-perception module</strong>, enhancing model performance on long-horizon and complex tasks.</li>
<li>Proposed and co-developed an <strong>active learning</strong> module for automated data labeling, significantly streamlining the data labeling pipeline.</li>
</ul>
<b></b>
<!-- <a href="https://arxiv.org/abs/" target="_blank">[Coming Soon]</a> -->
<!-- <p>This is an ongoing survey paper with awesome list repo.</p> -->
</p>
</div>
</div>
</article>
</ul>
<ul>
<li><strong>ProGEM: Probing Generalizable and Efficient Feature Fields Distillation for Multi-Task Robotic Manipulation</strong>:<br>
<i>Zhen Chen, Robin Ananda, Binxu Wu, <b>Haocheng Li</b>, Weize Li, Qi Lu, Qianyao Xu, Jiaying Li, Yingqing Xu</i><br>
<i>Under submission <br>
[Jul 2024 -- Nov 2024]</i><br>
</p>
<article class="columns">
<div class="column is-5">
<figure class="image">
<img src="images/publications/ProGEM_main.png">
</figure>
<figure class="image">
<img src="images/publications/ProGEM.gif">
</figure>
<br>
</div>
<div class="column">
<div class="content">
<ul>
<li>Used <strong>Stable Diffusion</strong> as a vision foundation model to extract semantic information from the scene.</li>
<li>Fused the semantic embeddings with <strong>RGB</strong> observations to build a semantically enriched <strong>3D point cloud</strong> (distilled feature field) of the scene.</li>
<li>Voxelized the scene and used a learned <strong>selection network</strong> to keep only the <strong>top-k occupied voxels</strong> from the voxel grid, so the policy reasons over a small set of informative voxels instead of the full grid — improving overall <strong>inference speed</strong>.</li>
</ul>
</div>
</div>
</article>
</ul>
<!--Academic Projects-->
<h2 id="Academic Projects">
Academic Projects
<span style="font-size: 1rem;margin-left: 1rem;position: relative;bottom: .2rem;">
</span>
</h2>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/publications/gridgame.gif">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b>Simulated Annealing AI Agent</b><br>
<a href="https://github.com/AdamLi111/Simulated_Annealing_Agent" target="_blank">[source code]</a>
</p>
<ul>
<li>Developed <b>simulated annealing AI agent</b> for constrained graph coloring, minimizing shapes while satisfying adjacency constraints</li>
<li>Implemented <b>adaptive optimization</b> with intelligent shape selection, problem area targeting, and dynamic cooling schedules</li>
<li>Engineered <b>efficient grid algorithms</b> using larger shapes first and offline state management for rapid solution evaluation</li>
<li>Applied <b>local search metaheuristics</b> to constraint satisfaction problems with applications in scheduling and resource allocation</li>
</ul>
</div>
</div>
</article>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/publications/escape_castle.gif">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b>Q-Learning Reinforcement Learning Agent</b><br>
<!-- <a href="https://github.com/haochenl/escape-the-castle-rl" target="_blank">[source code]</a><br> -->
</p>
<ul>
<li>Implemented <b>Q-learning agent</b> for grid-based MDP game with probabilistic transitions and combat mechanics</li>
<li>Developed <b>model-based probability estimation</b> from random gameplay to learn guard strength characteristics</li>
<li>Applied <b>epsilon-greedy exploration</b> with adaptive learning rates to train agent across 1M episodes</li>
<li>Built <b>temporal difference learner</b> that successfully navigates 5×5 castle grid while managing health and guard encounters</li>
</ul>
</div>
</div>
</article>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/publications/training_losses.png">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b>Deep Learning Image Classification with PyTorch</b><br>
<!-- <a href="https://github.com/haochenl/fashion-mnist-classifier" target="_blank">[source code]</a><br> -->
</p>
<ul>
<li>Designed and trained <b>CNN and feedforward neural networks</b> achieving 90%+ accuracy on Fashion-MNIST dataset classification</li>
<li>Implemented <b>convolutional architectures</b> with 3 conv layers, max pooling, and dropout regularization, demonstrating CNNs' superior performance with 25% fewer parameters than FFN</li>
<li>Visualized <b>learned feature representations</b> through kernel analysis and feature map progression, revealing hierarchical pattern detection from edges to complex shapes</li>
<li>Analyzed model performance using <b>confusion matrices and classification metrics</b>, identifying misclassification patterns between visually similar clothing categories</li>
</ul>
</div>
</div>
</article>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/publications/plannersystem.png">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b>Weekly Planner System</b><br>
<b>Multi-user system with role-based permissions</b><br>
<strong>Project leader</strong><br>
<a href="https://github.com/AdamLi111/OOD_hw" target="_blank">[source code]</a>
</p>
<ul>
<li>Built a multi-user weekly planner system with role-based permissions for event creation and deletion, ensuring automatic and real-time schedule updates for all related users</li>
<li>Chose <strong>MVC Design Pattern</strong> for this project, ensuring <strong>object-oriented design</strong> principles</li>
<li>Implemented XML functionalities where users can upload or save their schedule with an XML file, significantly improving system's efficiency</li>
</ul>
</div>
</div>
</article>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/publications/tuition_project.png">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b>Statewide University Statistics Analysis</b><br>
<b>Data Science Semester Project</b><br>
<strong>Project leader</strong><br>
<a href="https://github.com/AdamLi111/statewide_university_stat_comparison" target="_blank">[source code]</a>
</p>
<ul>
<li>Conducted a data-driven analysis of private universities by <strong>scraping</strong> key variables from CollegeSimply to investigate the relationship between tuition fees and educational quality</li>
<li>Implemented interactive linear regression plot, where every point on the plot is clickable and demonstrates the educational quality of a university using radar chart</li>
</ul>
</div>
</div>
</article>
<article>
<div class="columns">
<p><strong>[more projects coming soon]</strong></p>
</div>
</article>
<br>
<!--Professional Experience-->
<h2 id="Professional Experience">Professional Experience</h2>
<ul>
<li><strong>Northeastern University, EDGE Lab, SPHERE Project</strong>
<ul>
<li>[Jun 2025 -- Dec 2025]<strong>Research Engineer</strong><br>
<br>
<p>
- LeArm 6-DOF Robotic Arm System
<ul>
<li>Developed USB HID communication protocols for precise <b>6-DOF robotic arm control</b></li>
<li>Created automated interaction systems enabling physical manipulation of diverse IoT interfaces including button pressing and dial rotation</li>
<li>Built <b>Flask web servers</b> and SSH-based remote execution for distributed robotic control</li>
<li>Implemented position calibration and <b>inverse kinematics</b> for repeatable automated testing sequences</li>
</ul>
- XY-Plotter Drawing/Interaction System
<ul>
<li>Interfaced with <b>GRBL-based</b> CoreXY systems through G-code protocols</li>
<li>Implemented safety features including bounds checking and collision avoidance</li>
<li>Created high-level command abstractions (click, swipe, draw) from low-level motion primitives</li>
<li>Developed <b>real-time position tracking</b>and error recovery for autonomous operation</li>
</ul>
- SPHERE Framework Integration
<ul>
<li>Architected handler classes following design patterns for the SPHERE device infrastructure</li>
<li>Implemented <b>asynchronous command execution</b> using ThreadPoolExecutor</li>
<li>Built standardized interfaces and YAML-based configuration for flexible deployment</li>
</ul>
</p>
</li>
</ul>
</li>
<ul>
<li><strong>EncoSmart Technology</strong>
<ul>
<li>[Jul 2024 -- Nov 2024]<strong>Intern in the Department of Embodied AI</strong><br>
<p>
<ul>
<li>
Participated in the full research cycle of ProGEM. Assigned as the 4th author of the paper.
</li>
<li>Assisted in deploying our research result on the compnay's robot in real-world settings</li>
</ul>
</p>
<li>[Jul 2024 -- Sep 2024]<strong>Network Intern</strong><br>
<ul>
<li>Analyzed open-source IoT frameworks, leading to the selection of an optimal solution for the company's third generation product.</li>
<li>Assisted in developing and debugging of back-end interface code using Django for preliminary IoT networks.</li>
</ul>
<li>[Nov 2024 -- Dec 2024]<strong>Intern in the Department of Testing</strong><br>
<ul>
<li>Evaluated robotic arm and gripper's grasping stability across various objects, durations, and environmental conditions.</li>
<li>Assisted in the testing of back-end interface code of the preliminary IoT networks.</li>
</ul>
</ul></li>
</ul>
<!--Academic Service
<h2 id="Service">Service</h2>
<p>
I served / was delegated as Reviewer for NeurIPS’23, CVPR’24, IJCV, ICLR'25.
</p> -->
<!-- Misc. -->
<h2 id="Misc">Misc.</h2>
<p>
Outside of work and research, I enjoy rock climbing🧗, playing billiards🎱, and fitness💪. I also love to go fishing🎣 and skiing🏂 on vacations. Sometimes when I have the eureka moment, I also love to try implementing my ideas. See my little side projects below:
</p>
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/misc/fishing.jpg">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b><strong>Fishing🎣</strong></b>
</p>
<p>
I'm a huge fishing enthusiast. It's my favorite way to fully reset between large research or professional projects. Check out the largest fish I've caught by now. (There will be larger fish in the future and I will update this photo for sure)
</p>
</div>
</div>
</article>
<p>
My side projects:
<article class="columns">
<div class="column is-3">
<figure class="image">
<img src="images/misc/MBTI.png">
</figure>
</div>
<div class="column">
<div class="content">
<p>
<b><strong>Fight of MBTI's</strong></b>
<p>An online card game I designed and implemented using Java. Currently supports up to 4 players. Players can play cards with different abilities to reduce enemies' health or defense. Last one survived wins the game.
</p>
</p>
</div>
</div>
</article>
<!-- <figure class="image">
<img src="images/misc/MBTI.png">
</figure> -->
</p>
<!-- To be add? -->
</div>
</div>
</div>
</div>
<!-- Footer -->
<footer class="container">
<br><hr>
<!-- counts ip -->
<div id="mapmyvisitors-widget" style="width:10%; margin: 0 auto; text-align: center;">
<script type="text/javascript" id="mmvst_globe" src="//mapmyvisitors.com/globe.js?d=afdJ46ACXFHtiC_wv5ZfMzpEBedGslwjhPprv68F7rw"></script>
</div>
<div class="row" style="text-align: center;">
© 2025 Haocheng Li<br>Last update: Oct. 2026
</div>
</footer>
</section>
<script>
var $hashList = $('.menu-list a'), offsetList, maxScrollHeight;
$('#sidebar').on('click', 'a', function(){
activate($(this))
});
$(window).on('resize', debounce(calculateBoundary, 300));
$(document).on('scroll', debounce(judgeScroll, 300));
calculateBoundary();
judgeScroll();
function togglePatents() {
var content = document.getElementById("patentsContent");
if (content.style.display === "none") {
content.style.display = "block";
} else {
content.style.display = "none";
}
}
function calculateBoundary() {
offsetList = $hashList.map(function(idx, ele){
return $(ele.hash).offset().top
});
maxScrollHeight = $(document).height() - $(window).height()
}
function judgeScroll() {
var tps = $("html").scrollTop()
? $("html").scrollTop()
: $("body").scrollTop(),
len = offsetList.length;
if(tps >= maxScrollHeight-10){
activate($hashList.eq(len-1));
return
}
for(var i=0; i<len; i++){
if(tps+50<offsetList[i]){
activate($hashList.eq(Math.max(0,i-1)));
return
}
}
}
function activate(ele){
$hashList.removeClass('is-active');
ele.addClass('is-active');
}
function debounce(fn, delay) {
var timeout = null;
return function () {
var args = arguments;
var context = this;
if (!timeout) {
timeout = setTimeout(function () {
timeout = 0;
return fn.apply(context, args);
}, delay);
}
};
}
</script>
</body>
</html>