Skip to content

Commit 96e3277

Browse files
jyao97jyao97
committed
added a sample research project to research
1 parent c9488d0 commit 96e3277

File tree

6 files changed

+132
-61
lines changed

6 files changed

+132
-61
lines changed

‎_data/projects.yaml

-47
This file was deleted.

‎_data/research.yaml

+21
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,21 @@
1+
- title: Explainable Relational Reasoning and Multi-Agent Interaction Modeling (Social & Physical)
2+
# subtitle: a subtitle
3+
# group: featured
4+
image: research_images/graph2.webp
5+
# link: https://github.com/
6+
description: |
7+
We investigate relational reasoning and interaction modeling in the context of the trajectory prediction task, which aims to generate accurate, diverse future trajectory hypotheses or state sequences based on historical observations.
8+
Our research introduced the first unified relational reasoning toolbox that systematically infers the underlying relations/interactions between entities at different scales (e.g., pairwise, group-wise) and different abstraction levels (e.g., multiplex) by learning dynamic latent interaction graphs and hypergraphs from observable states (e.g., positions) in an unsupervised manner.
9+
The learned latent graphs are explainable and generalizable, significantly improving the performance of downstream tasks, including prediction, sequential decision making, and control.
10+
We also proposed a physics-guided relational learning approach for physical dynamics modeling.
11+
12+
**Related Publications\:** <br>
13+
1. [EvolveGraph: Multi-Agent Trajectory Prediction with Dynamic Relational Reasoning](https://proceedings.neurips.cc/paper/2020/hash/e4d8163c7a068b65a64c89bd745ec360-Abstract.html), NeurIPS 2020. <br>
14+
2. [RAIN: Reinforced Hybrid Attention Inference Network for Motion Forecasting](https://openaccess.thecvf.com/content/ICCV2021/html/Li_RAIN_Reinforced_Hybrid_Attention_Inference_Network_for_Motion_Forecasting_ICCV_2021_paper.html), ICCV 2021. <br>
15+
3. [Interaction Modeling with Multiplex Attention](https://proceedings.neurips.cc/paper_files/paper/2022/hash/7e6361a5d73a8fab093dd8453e0b106f-Abstract-Conference.html), NeurIPS 2022. <br>
16+
4. [Learning Physical Dynamics with Subequivariant Graph Neural Networks](https://proceedings.neurips.cc/paper_files/paper/2022/hash/a845fdc3f87751710218718adb634fe7-Abstract-Conference.html), NeurIPS 2022. <br>
17+
5. [Grouptron: Dynamic Multi-Scale Graph Convolutional Networks for Group-Aware Dense Crowd Trajectory Forecasting](https://arxiv.org/abs/2109.14128), ICRA 2022. <br>
18+
6. [Multi-Agent Dynamic Relational Reasoning for Social Robot Navigation](https://arxiv.org/abs/2401.12275), submitted to IEEE Transactions on Robotics (T-RO), under review.
19+
# repo: greenelab/lab-website-template
20+
tags:
21+
- Graph Neural Networks

‎_includes/card_research.html

+50
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,50 @@
1+
{{ " " }}
2+
<div class="card" data-style="{{ include.style }}" style="width: 100%; ">
3+
<div class="card-img" style="aspect-ratio: 5.0; overflow: hidden;">
4+
<a
5+
{% if include.link %}
6+
href="{{ include.link | relative_url }}"
7+
{% endif %}
8+
aria-label="{{ include.title | default: "card link" }}"
9+
class="card-image"
10+
>
11+
<img
12+
src="{{ include.image | relative_url }}"
13+
alt="{{ include.title | default: "card image" }}"
14+
loading="lazy"
15+
style="width: 100%; height: auto; object-fit: contain;"
16+
{% include fallback.html %}
17+
>
18+
</a>
19+
</div>
20+
21+
<div class="card-text">
22+
{% if include.title %}
23+
<a
24+
{% if include.link %}
25+
href="{{ include.link | relative_url }}
26+
{% endif %}
27+
{% if include.tooltip %}
28+
data-tooltip="{{ include.tooltip }}"
29+
{% endif %}
30+
class="card-title"
31+
>
32+
{{ include.title }}
33+
</a>
34+
{% endif %}
35+
36+
{% if include.subtitle %}
37+
<span class="card-subtitle">{{ include.subtitle }}</span>
38+
{% endif %}
39+
40+
{% if include.description %}
41+
<p>
42+
{{ include.description | markdownify | remove: "<p>" | remove: "</p>" }}
43+
</p>
44+
{% endif %}
45+
46+
{% if include.tags or include.repo %}
47+
{% include tags.html tags=include.tags repo=include.repo %}
48+
{% endif %}
49+
</div>
50+
</div>

‎_includes/list_no_group.html

+59
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,59 @@
1+
{% assign emptyarray = "" | split: "," %}
2+
{% assign data = site.data[include.data]
3+
| default: site[include.data]
4+
| default: emptyarray
5+
| data_filter: include.filters
6+
%}
7+
8+
{% assign years = data
9+
| group_by_exp: "d", "d.date | date: '%Y'"
10+
| sort: "name"
11+
| reverse
12+
%}
13+
14+
{% for year in years %}
15+
{% assign data = year.items %}
16+
17+
{% if years.size > 1 %}
18+
{{--}}<h3 id="{{ year.name }}">{{ year.name }}</h3>
19+
{% assign data = data | sort: "date" | reverse %}
20+
{% endif %}
21+
22+
{% for d in data %}
23+
{% assign style = d.style | default: include.style %}
24+
25+
{%
26+
include {{ include.component | append: ".html" }}
27+
affiliation=d.affiliation
28+
author=d.author
29+
authors=d.authors
30+
buttons=d.buttons
31+
caption=d.caption
32+
content=d.content
33+
date=d.date
34+
description=d.description
35+
excerpt=d.excerpt
36+
height=d.height
37+
icon=d.icon
38+
id=d.id
39+
image=d.image
40+
last_modified_at=d.last_modified_at
41+
link=d.link
42+
lookup=d.lookup
43+
name=d.name
44+
publisher=d.publisher
45+
repo=d.repo
46+
role=d.role
47+
slug=d.slug
48+
style=style
49+
subtitle=d.subtitle
50+
tags=d.tags
51+
text=d.text
52+
title=d.title
53+
tooltip=d.tooltip
54+
type=d.type
55+
url=d.url
56+
width=d.width
57+
%}
58+
{% endfor %}
59+
{% endfor %}

‎research/index.md

+2-14
Original file line numberDiff line numberDiff line change
@@ -10,17 +10,5 @@ nav:
1010

1111
The ultimate goal of our research is to build **trustworthy**, **interactive**, and **human-centered** autonomous agents that can perceive, understand, and reason about the physical world; safely interact and collaborate with humans and other agents, and clearly explain their behaviors to build trust with humans so that they can benefit society in daily lives. To achieve this goal, we have been pursuing interdisciplinary research and unifying the techniques and tools from robotics, machine learning, reinforcement learning, explainable AI, control theory, optimization, and computer vision.
1212

13-
14-
{% include search-info.html %}
15-
16-
{% include section.html %}
17-
18-
## Featured
19-
20-
{% include list.html component="card" data="projects" filters="group: featured" %}
21-
22-
{% include section.html %}
23-
24-
## More
25-
26-
{% include list.html component="card" data="projects" filters="group: " style="small" %}
13+
<!-- {% include list_no_group.html component="card" data="projects" filters="group: featured" %} -->
14+
{% include list_no_group.html component="card_research" data="research"%}

‎research_images/graph2.webp

841 KB
Binary file not shown.

0 commit comments

Comments
 (0)