CES 2016: Audi, Daimler, BMW and More (part 6)

TL;DR

Deep learning on NVIDIA GPUs for self-driving cars surpasses human recognition in road sign detection.

Transcript

so so that deep that network called dr net is running on currently a tight necks and and if somebody can later bring me a tight neck so i could hold up what a Titan ex looks like a tight next is basically invidious highest performance GPU that's used in desktop pcs running on a tight necks running on a tight next just now the MV drive net can achie... Read More

Key Insights

• ✋ Titan X GPUs enable high-performance deep learning for self-driving networks.
• 🤘 Audi engineers achieve superior road sign recognition using NVIDIA's platform.
• 😨 Companies like Daimler and BMW leverage deep learning for self-driving car advancements.
• 🪛 Drive PX2 combines supercomputing capabilities with real-time computer graphics for self-driving systems.
• ✊ DriveWorks, NVIDIA's operating system, algorithms, and streaming pipelines, powers deep learning in self-driving systems.
• ❓ The future includes training for recognizing circumstances beyond object detection.
• 🐕‍🦺 Deep learning applications extend to various industries like robotics and manufacturing, enhancing efficiency and smart services.

Questions & Answers

Q: What GPU is being used for deep learning in self-driving car development?

The Titan X GPU is utilized for deep learning in self-driving car development, achieving high performance of 50 frames per second.

Q: How does the network developed by Audi engineers compare to human recognition?

Audi engineers trained a network that surpassed human recognition in road sign detection using NVIDIA's platform for deep learning.

Q: Which companies are using NVIDIA's platform for deep learning besides Audi?

Companies like Daimler, BMW, preferred networks, and Ford are leveraging NVIDIA's platform for advancements in self-driving cars and AI applications.

Q: What challenges are faced in processing sensor information for self-driving cars?

Processing vast amounts of sensor information poses a system software challenge, requiring advanced algorithms and engineering efforts to run efficiently on platforms like Drive PX2.

Summary & Key Takeaways

• Deep learning on Titan X GPUs achieves 50 frames per second for self-driving network development.

• Audi engineers utilize NVIDIA platform to train networks surpassing human capability in road sign recognition.

• Daimler, BMW, and other companies leverage NVIDIA deep learning for self-driving car advancements.