У нас вы можете посмотреть бесплатно Autonomous Driving with Bidirectional Negotiation and Wrong Side Traffic Negotiation with Jan Kuenne или скачать в максимальном доступном качестве, видео которое было загружено на ютуб. Для загрузки выберите вариант из формы ниже:
Если кнопки скачивания не
загрузились
НАЖМИТЕ ЗДЕСЬ или обновите страницу
Если возникают проблемы со скачиванием видео, пожалуйста напишите в поддержку по адресу внизу
страницы.
Спасибо за использование сервиса ClipSaver.ru
On 13th March Jan Kuenne visited us for a ride in our #autonomousvehicles and celebrate Holi with us. We conducted two demos with him, showcasing the key research frontiers that Swaayatt Robots (स्वायत्त रोबोट्स) is working on towards achieving Level-5 #autonomousdriving. Demo-1 involved our navigation on a single lane road, with bidirectional traffic. The road is wide enough to allow only one vehicle passage at a time -- for any two vehicles approaching each other head-on, both will have to yield, and shift off-roads. Over the past 1.5 years, we have made significant progress towards learning such navigation policies, i.e., bidirectional traffic negotiation on single lane roads. Furthermore, the task was made even more challenging with incoming traffic approaching from the wrong side of the road, not abiding by the driving rules, making our autonomous vehicle use onboard intelligence to also negotiate from the wrong side -- and not being restricted to the driving laws and get stuck -- demonstrating (time-) optimal decision making and motion planning for complex sequentially-stochastic navigation tasks. Our (#reinforcementlearning-) agents, i.e., the motion planning and decision making algorithmic framework successfully executed this task. Demo-2 was high-speed maneuvering among traffic cones, avoiding them at extremely high speeds (near-drift), with limited visibility at night. This will be posted later this week. There is much more coming soon! Going forward, we will keep scaling our algorithmic frameworks and decision making agents that are targeted towards Level-5 autonomous driving. #deeplearning