1+ <!DOCTYPE html>
2+ < html >
3+
4+ < head >
5+ < meta charset ="utf-8 ">
6+ < meta name ="description " content ="Soccer behavior learning in parallel environments ">
7+ < meta name ="keywords " content ="Reinforcement Learning, Robotics, Simulation, PPO ">
8+ < meta name ="viewport " content ="width=device-width, initial-scale=1 ">
9+ < title > DNT-RL: Soccer Behavior Learning in Parallel Environments</ title >
10+
11+ < link href ="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro " rel ="stylesheet ">
12+
13+ < link rel ="stylesheet " href ="./static/css/bulma.min.css ">
14+ < link rel ="stylesheet " href ="./static/css/bulma-carousel.min.css ">
15+ < link rel ="stylesheet " href ="./static/css/bulma-slider.min.css ">
16+ < link rel ="stylesheet " href ="./static/css/fontawesome.all.min.css ">
17+ < link rel ="stylesheet " href ="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css ">
18+ < link rel ="stylesheet " href ="./static/css/index.css ">
19+ < link rel ="icon " href ="./static/images/favicon.svg ">
20+
21+ < script src ="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js "> </ script >
22+ < script defer src ="./static/js/fontawesome.all.min.js "> </ script >
23+ < script src ="./static/js/bulma-carousel.min.js "> </ script >
24+ < script src ="./static/js/bulma-slider.min.js "> </ script >
25+ </ head >
26+
27+ < body >
28+
29+ < nav class ="navbar " role ="navigation " aria-label ="main navigation ">
30+ < div class ="navbar-brand ">
31+ < a role ="button " class ="navbar-burger " aria-label ="menu " aria-expanded ="false ">
32+ < span aria-hidden ="true "> </ span >
33+ < span aria-hidden ="true "> </ span >
34+ < span aria-hidden ="true "> </ span >
35+ </ a >
36+ </ div >
37+ < div class ="navbar-menu ">
38+ < div class ="navbar-start " style ="flex-grow: 1; justify-content: center; ">
39+ </ div >
40+ </ div >
41+ </ nav >
42+
43+
44+ < section class ="hero ">
45+ < div class ="hero-body ">
46+ < div class ="container is-max-desktop ">
47+ < div class ="columns is-centered ">
48+ < div class ="column has-text-centered ">
49+ < h1 class ="title is-1 publication-title "> DNT-RL: Soccer Behavior Learning in Parallel Environments</ h1 >
50+ < div class ="is-size-5 publication-authors ">
51+ < span class ="author-block ">
52+ Macha Meijer< sup > 1,2</ sup > ,</ span >
53+ < span class ="author-block ">
54+ Gijs de Jong< sup > 1,2</ sup > ,</ span >
55+ < span class ="author-block ">
56+ Joshua Rosenthal< sup > 1,2</ sup >
57+ </ span >
58+ </ div >
59+
60+ < div class ="is-size-5 publication-authors ">
61+ < span class ="author-block "> < sup > 1</ sup > University of Amsterdam,</ span >
62+ < span class ="author-block "> < sup > 2</ sup > Dutch Nao Team</ span >
63+ </ div >
64+
65+ < div class ="column has-text-centered ">
66+ < div class ="publication-links ">
67+ <!-- PDF Link. -->
68+ < span class ="link-block ">
69+ < a href ="TODO " class ="external-link button is-normal is-rounded is-dark ">
70+ < span class ="icon ">
71+ < i class ="fas fa-file-pdf "> </ i >
72+ </ span >
73+ < span > Paper</ span >
74+ </ a >
75+ </ span >
76+ < span class ="link-block ">
77+ < a href ="TODO " class ="external-link button is-normal is-rounded is-dark ">
78+ < span class ="icon ">
79+ < i class ="ai ai-arxiv "> </ i >
80+ </ span >
81+ < span > arXiv</ span >
82+ </ a >
83+ </ span >
84+ <!-- Video Link. TODO-->
85+ < span class ="link-block ">
86+ < a href ="TODO " class ="external-link button is-normal is-rounded is-dark ">
87+ < span class ="icon ">
88+ < i class ="fab fa-youtube "> </ i >
89+ </ span >
90+ < span > Video</ span >
91+ </ a >
92+ </ span >
93+ <!-- Code Link. -->
94+ < span class ="link-block ">
95+ < a href ="https://github.com/IntelligentRoboticsLab/dnt-rl "
96+ class ="external-link button is-normal is-rounded is-dark ">
97+ < span class ="icon ">
98+ < i class ="fab fa-github "> </ i >
99+ </ span >
100+ < span > Code</ span >
101+ </ a >
102+ </ span >
103+ <!-- Simulation Link. -->
104+ < span class ="link-block ">
105+ < a href ="https://github.com/oxkitsune/striker/ "
106+ class ="external-link button is-normal is-rounded is-dark ">
107+ < span class ="icon ">
108+ < i class ="fa fa-robot "> </ i >
109+ </ span >
110+ < span > Physics Simulator</ span >
111+ </ a >
112+ </ div >
113+ </ div >
114+ </ div >
115+ </ div >
116+ </ div >
117+ </ div >
118+ </ section >
119+
120+ < section class ="hero teaser ">
121+ < div class ="container is-max-desktop ">
122+ < div class ="hero-body ">
123+ < video id ="teaser " autoplay muted loop playsinline height ="100% ">
124+ < source src ="./static/videos/striker.webm " type ="video/webm ">
125+ </ video >
126+ < h2 class ="subtitle has-text-centered ">
127+ < span class ="ddnt-rl "> DNT-RL</ span > learning a walk to point policy with 64 parallel agents.
128+ </ h2 >
129+ </ div >
130+ </ div >
131+ </ section >
132+
133+
134+ < section class ="hero is-light is-small ">
135+ < div class ="hero-body ">
136+ < div class ="container ">
137+ < div id ="results-carousel " class ="carousel results-carousel ">
138+ < div class ="item item-steve ">
139+ < video poster ="" id ="steve " autoplay controls muted loop playsinline height ="100% ">
140+ < source src ="./static/videos/striker.webm " type ="video/webm ">
141+ </ video >
142+ </ div >
143+ <!-- <div class="item item-chair-tp">
144+ <video poster="" id="chair-tp" autoplay controls muted loop playsinline height="100%">
145+ <source src="./static/videos/chair-tp.mp4" type="video/mp4">
146+ </video>
147+ </div>
148+ <div class="item item-shiba">
149+ <video poster="" id="shiba" autoplay controls muted loop playsinline height="100%">
150+ <source src="./static/videos/shiba.mp4" type="video/mp4">
151+ </video>
152+ </div>
153+ <div class="item item-fullbody">
154+ <video poster="" id="fullbody" autoplay controls muted loop playsinline height="100%">
155+ <source src="./static/videos/fullbody.mp4" type="video/mp4">
156+ </video>
157+ </div>
158+ <div class="item item-blueshirt">
159+ <video poster="" id="blueshirt" autoplay controls muted loop playsinline height="100%">
160+ <source src="./static/videos/blueshirt.mp4" type="video/mp4">
161+ </video>
162+ </div>
163+ <div class="item item-mask">
164+ <video poster="" id="mask" autoplay controls muted loop playsinline height="100%">
165+ <source src="./static/videos/mask.mp4" type="video/mp4">
166+ </video>
167+ </div>
168+ <div class="item item-coffee">
169+ <video poster="" id="coffee" autoplay controls muted loop playsinline height="100%">
170+ <source src="./static/videos/coffee.mp4" type="video/mp4">
171+ </video>
172+ </div>
173+ <div class="item item-toby">
174+ <video poster="" id="toby" autoplay controls muted loop playsinline height="100%">
175+ <source src="./static/videos/toby2.mp4" type="video/mp4">
176+ </video>
177+ </div> -->
178+ </ div >
179+ </ div >
180+ </ div >
181+ </ section >
182+
183+
184+ < section class ="section ">
185+ < div class ="container is-max-desktop ">
186+ <!-- Abstract. -->
187+ < div class ="columns is-centered has-text-centered ">
188+ < div class ="column is-four-fifths ">
189+ < h2 class ="title is-3 "> Abstract</ h2 >
190+ < div class ="content has-text-justified ">
191+ < p >
192+ Reinforcement Learning (RL) has shown significant potential in the RoboCup Soccer leagues, for example in
193+ the SPL.
194+ </ p >
195+ < p >
196+ Unlike tranditional hand-crafted behaviors, the usage of RL enables the development a more flexible
197+ behavior engine, allowing for a more dynamic gameplay.
198+ However, training RL behaviors is resource-intensive, both in terms of training time and the development
199+ time of each behavior.
200+ </ p >
201+ < p >
202+ In this paper, we introduce t < span class ="ddnt-rl "> DNT-RL</ span > , a framework designed for the efficient
203+ training of 2D RL behaviors
204+ using parallel training with multiple environments. The framework integrates a physics-based simulation
205+ with a standardized way of training behaviors, supporting parallelized and GPU-accelerated training to
206+ reduce the training time.
207+ We show the effectiveness of DNT-RL through several behaviors that are trained using the framework, which
208+ have been succesfully deployed in RoboCup tournaments. These results show the potential of the framework
209+ for development of scalable and usable RL-based behavior training.
210+ </ p >
211+ </ div >
212+ </ div >
213+ </ div >
214+ <!--/ Abstract. -->
215+
216+ <!-- Paper video. -->
217+ <!-- <div class="columns is-centered has-text-centered">
218+ <div class="column is-four-fifths">
219+ <h2 class="title is-3">Video</h2>
220+ <div class="publication-video">
221+ <iframe src="https://www.youtube.com/embed/MrKrnHhk8IA?rel=0&showinfo=0" frameborder="0"
222+ allow="autoplay; encrypted-media" allowfullscreen></iframe>
223+ </div>
224+ </div>
225+ </div> -->
226+ <!--/ Paper video. -->
227+ </ div >
228+ </ section >
229+
230+
231+ < section class ="section ">
232+ < div class ="container is-max-desktop ">
233+
234+
235+ <!-- Concurrent Work. -->
236+ < div class ="columns is-centered ">
237+ < div class ="column is-full-width ">
238+ < h2 class ="title is-3 "> Related Links</ h2 >
239+
240+ < div class ="content has-text-justified ">
241+ < p >
242+ There's a lot of excellent work that was introduced around the same time as ours.
243+ </ p >
244+ < p >
245+ < a href ="https://arxiv.org/abs/2412.09417v2 "> Reinforcement Learning Within the Classical Robotics Stack: A
246+ Case Study in Robot Soccer</ a > introduces
247+ an
248+ idea similar to our classic robotics stack enhanced with RL behavior.
249+ </ p >
250+ </ div >
251+ </ div >
252+ </ div >
253+ <!--/ Concurrent Work. -->
254+
255+ </ div >
256+ </ section >
257+
258+
259+ < section class ="section " id ="BibTeX ">
260+ < div class ="container is-max-desktop content ">
261+ < h2 class ="title "> BibTeX</ h2 >
262+ < pre > < code > @article{meijer2025dntrl,
263+ author = {Meijer, Macha and de Jong, Gijs and Rosenthal, Joshua},
264+ title = {DNT-RL: Soccer Behavior Learning in Parallel Environments},
265+ journal = {% TODO}
266+ year = {2025},
267+ }</ code > </ pre >
268+ </ div >
269+ </ section >
270+
271+
272+ < footer class ="footer ">
273+ < div class ="container ">
274+ < div class ="columns is-centered ">
275+ < div class ="column ">
276+ < div class ="content has-text-centered ">
277+ < p >
278+ Website template borrowed from < a href ="https://github.com/nerfies/nerfies.github.io "> NeRFies</ a > and < a
279+ href ="https://genesis-embodied-ai.github.io "> Genesis</ a > .
280+ </ p >
281+ </ div >
282+ </ div >
283+ </ div >
284+ </ div >
285+ </ footer >
286+
287+ </ body >
288+
289+ </ html >
0 commit comments