![Duckietown](/img/default-banner.jpg)
- Видео 90
- Просмотров 42 203
Duckietown
США
Добавлен 15 авг 2018
Starting from the Massachusetts Institute of Technology (MIT), Duckietown is a platform for delivering cutting-edge robotics and AI learning experiences tailored to all skill levels.
We offer teaching resources to instructors, hands-on activities to learners, an accessible research platform to researchers, and a state-of-the-art ecosystem for professional training.
We believe the world needs more talent to shape the next generations of robot autonomy, and we know that talent is you.
We offer teaching resources to instructors, hands-on activities to learners, an accessible research platform to researchers, and a state-of-the-art ecosystem for professional training.
We believe the world needs more talent to shape the next generations of robot autonomy, and we know that talent is you.
Duckiedrone (DD24) - Assembly teaser
Learn more about the Duckiedrone (DD24) and start teaching and learning autonomy through flight at: docs.duckietown.com/daffy/opmanual-dd24/intro.html
Просмотров: 98
Видео
Brain Computer Interfaces (BCI) initiative - 2024 MIT Hackathon - winning team experience
Просмотров 1383 месяца назад
For more information on the BCI initiative check out: bci-i.github.io/
Duckiebot lane following pipeline
Просмотров 683Год назад
Lane following pipeline 1. color correction 2. detect line segments 3. extract pose estimate from each segment leveraging Duckietown appearance specifications 4. run a histogram filter in real-time, on a Raspberry Pi, and elect the most "voted" pose as the Duckiebot's estimate 5. (control motors) 6. repeat Do it yourself at duckietown.com
The AI Driving Olympics
Просмотров 267Год назад
Prof. Liam Paull explains the AI Driving Olympics in the occasion of the third edition that took place at ICRA 2019 in Montréal.
Duckietown City Rescue Demo
Просмотров 176Год назад
Student project ETH Zurich, 2018 A Duckietown Autolab detects crashed Duckiebots and puts them back on track. Demonstration of real-time localization and i2v communication.
Duckietown - Supervised learning demo
Просмотров 171Год назад
Lane following with supervised learning DB17
Duckietown - Golden Duckie Award - the Green Station
Просмотров 77Год назад
Design and Prototype by Andrea F. Daniele
Duckietown - Improving lane detection and control
Просмотров 127Год назад
Project by: Marco Stalder, Simon Muntwiler, Anna Dai, Manuel Breitenstein, Andreas Aumiller, Miguel De La Iglesia at ETHZ 2017
Illumination invariance in Duckietown - the "Anti-Instagram filter" node
Просмотров 40Год назад
By Christoph Zuidema, Milan Schilling, Shengjie Lin, and David Yunis at ETHZ, 2017
Duckietown Autolab Localization v. 0.5 - AI Driving Olympics
Просмотров 106Год назад
Duckietowns augmented to Autolabs with watchtowers, detect Duckiebots in town and track their pose over time. The data is then used to quantify agent performance in specific tasks. Learn all about the AI-DO on www.duckietown.org/research/ai-driving-olympics
Duckietown - System Identification Demo
Просмотров 131Год назад
Using camera based detection of the calibration pattern and open loop control to identify the motion's model parameters of the Duckiebot
Duckietown DB17 Duckiebots Formation Keeping
Просмотров 53Год назад
DB17 Duckiebots perform computer-vision-based traffic management by detecting the circle pattern and using PID to maintain a fixed relative pose. Video by students of the 2017 edition of AMoD at ETH Zürich
Duckietown: Megacity Demo
Просмотров 109Год назад
"Megacity" demo in the ETHZ Autolab, where multiple Duckiebots navigate a Duckietown Autolab.
Sample color detector output
Просмотров 18Год назад
A simple duckie-detector based on color detection.
Duckietown (DB17) Intersection Navigation and Coordination
Просмотров 57Год назад
Duckietown (DB17) Intersection Navigation and Coordination
DBv2 lane following with line sensors only
Просмотров 14Год назад
DBv2 lane following with line sensors only
Duckietown at night - ML based lane-following
Просмотров 16Год назад
Duckietown at night - ML based lane-following
Duckietown Autolab: "auto-charging" v. 0.1 demo
Просмотров 38Год назад
Duckietown Autolab: "auto-charging" v. 0.1 demo
Click-and-move Duckiebot with front facing arm
Просмотров 109Год назад
Click-and-move Duckiebot with front facing arm
Development of an Ackerman Steering Duckietown vehicle
Просмотров 2272 года назад
Development of an Ackerman Steering Duckietown vehicle
awesome tie
Short and good explanation! Thanks a lot!
Quack!
Thanks a lot, great tutorial!
Great tie! 😄
thanks ❤
that was great explanation in easy way
Takeawahs: 1. Motion planning can be formulated as an optimal control problem. In the case of autonomous cars, we want to formlulate a command profile that creates a desireable motion within a set of cost parameters, such as comfort, fuel efficiency, path length. Commands are such as steering, gas, and brakes, or idealized, such as vector, speed, and acceleration.j 2. This approach is tractable only in simple cases. Next video: how to we simplify the problem so it is tractable?
Very good explanation, thanks !
This is so awesome. The new design looks so amazing.
Niceee
Ah! Love the Trailer at the back, waiting for the 4x4 AWD duckiebot one day!
Very informative video, Thank you 👏
Whyn you guys don't post videos on regular basis!
Duckietown! Learning on EDX.
thank you👍