How Deep Learning will Enable Self-Driving Cars

Deep learning refers to algorithms—step-by-step data-crunching recipes—for teaching machines to understand unstructured data, such as images, speech and video.
With deep learning, a neural network learns many levels of abstraction. This lets a machine learn what computer scientists call a “hierarchical representation,” and gets smarter the more information is fed into the system. The result is that a deep neural network for driving applications can understand the subtle nuances of what is happening around the vehicle. It can discern an ambulance from a delivery truck, determine whether a parked car is vacant or the door is opening as a passenger emerges, and also detect occluded objects, such as a pedestrian partially blocked by a parked car.
GPUs are ideal for training the neural net, and for running the model inside the vehicle in real time. They can cut the time that it takes to train these neural networks to just days from a year or more, as the GPU is a massively parallel processor. Once a system is “trained,” that learning can be used in applications for self-driving cars.

Speaker: Mike Houston

Michael Houston is a Distinguished Engineer in the Deep Learning business at NVIDIA, which is paving the way to self-driving cars, computers that detect tumors, and real-time speech translation. Michael has more than 12 years of experience in the computer graphics industry. He joined NVIDIA in 2012, after working at AMD as a Systems Architect. Michael earned a PhD in computer science from Stanford University.

Agenda:

6:30-7:00 PM- Networking, pizza, and refreshments

7:00- 8:00 PM- Deep Learning and Self-Driving Cars

8:00-8:15 PM- Q&A

8:15-8:30 PM- Networking

Thanks to Nvidia for organizing and hosting this event! As always, 100% of proceeds go to benefit San Jose State University's Formula SAE racing team. (I am not an alumnus of SJSU. -- Alison)