| Session Description | |
|---|---|
| Title: | CUDA FOR MACHINE VISION AND DRIVER ASSISTANCE: GPU Computing in Tomorrow's Automobiles |
| Track: | Technical Conference |
| Schedule: |
August 27, 2008 10:25 AM - 11:05 AM Hilton Hotel/San Carlos Room 1 |
|
In cooperation with engineers at the BMW Group Technology Office in Palo Alto, and NVIDIA Corporation in Santa Clara, we are conducting research on automotive computing using a programmable graphics processing unit (GPU). This talk will present our ideas on computing in tomorrow's automobiles and explain why GPUs are a good fit for realizing these future goals. Having the low-cost, high-performance GPU as an embedded processor in a car would provide ample compute power for many compelling applications. For instance, in our project we are investigating the implementation of computer vision applications on a GPU to provide the driver with an enhanced awareness of his or her surroundings. This presentation will provide examples of such data-parallel automotive tasks and discuss what is good and bad about GPUs for performing these tasks. |
|
| Speakers |
|---|
|
Pinar Muyan-Ozcelik (UC Davis/BMW)
Pinar Muyan-Ozcelik is a PhD student in Computer Science at UC Davis. She is interested in GPU Computing and has been conducting research on GPGPU in medical imaging and computer vision under the supervision of Dr. John Owens. Pinar received her MSc degree in Computer Science from the University of British Columbia, Vancouver, Canada and BSc degree in Computer Engineering from Ege University, Izmir, Turkey. In terms of professional experience, she has worked as a System Analyst in SanMar, Canada, and as a Software Developer in ImageMax, Indianapolis, IN. |
| Vladimir Glavtchev (BMW) |
| close |