Timezone: »

Accelerating Deep Neural Networks on Mobile Processor with Embedded Programmable Logic
Eugenio Culurciello · Aysegul Dundar · Jonghoon Jin · Vinayak Gokhale · Berin Martini

Fri Dec 06 07:00 PM -- 11:59 PM (PST) @ Tahoe A+B, Harrah’s Special Events Center 2nd Floor
Event URL: Some old videos can be found here: http://www.neuflow.org/category/videos/ »

We present a live demonstration of a mobile platform aimed at accelerating deep convolutional neural networks (DCNNs). DCNNs is a powerful way to categorize images. They have achieved state of the art performance in many visual classification benchmarks and have won many competitions. However, their computational costs prevent them from being deployed for real-time applications. We implemented a hardware accelerator on the Xilinx Zynq SoC that can run DCNNs in real-time. The platform consists of a FPGA (PL) and two ARM Cortex-A9 cores (PS). The PL and PS share the same DDR3 memory which allows us to achieve a very high throughput when transferring data between software and co-processor. We will demonstrate live applications of DCNNs on our hardware.

Author Information

Eugenio Culurciello (FWDNXT)
Aysegul Dundar (Purdue University)
Jonghoon Jin (Purdue University)
Vinayak Gokhale (Purdue University)
Berin Martini (Purdue University)

More from the Same Authors