Тёмный
No video :(

Differentiable Physics (for Deep Learning), Overview Talk by Nils Thuerey 

Nils Thuerey
Подписаться 2,6 тыс.
Просмотров 8 тыс.
50% 1

Опубликовано:

 

5 сен 2024

Поделиться:

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист
Посмотреть позже
Комментарии : 7   
@nitefather
@nitefather 4 года назад
I will be starting a PhD thesis in the use of deep learning for spray and atomization simulations, The work by your group has been a great source of inspiration, thank you for sharing!
@HerrWortel
@HerrWortel 3 года назад
Do you by any chance employ Graph Neural Nets for such simulations?
@nitefather
@nitefather 3 года назад
@@HerrWortel hello! Our work is recently starting, for the moment we haven't experimented with GNNs but for sure they offer great potential, I suggest the work done peter battaglia and sanchez-gonzalez, great stuff.
@annekedebruyn7797
@annekedebruyn7797 4 года назад
With these amount of speed ups, you guys are literally going to be the people that will invent proper real time interactive simulations that don''t look like early 90''s water sims.
@nikronic
@nikronic 4 года назад
That was an amazing work. This idea can be used in almost all areas that try to simulate a behavior through time (or other variables).
@varshneydevansh
@varshneydevansh 11 месяцев назад
@kvasios
@kvasios Год назад
39:26 amen to that!
Далее
Brax: A New Differentiable Physics Engine
50:20
Просмотров 4,5 тыс.
AI for physics & physics for AI
49:04
Просмотров 27 тыс.