Title :
Accelerating real-time embedded scene labeling with convolutional networks
Author :
Cavigelli, Lukas ; Magno, Michele ; Benini, Luca
Author_Institution :
Integrated Syst. Lab., ETH Zurich, Zurich, Switzerland
Abstract :
Today there is a clear trend towards deploying advanced computer vision (CV) systems in a growing number of application scenarios with strong real-time and power constraints. Brain-inspired algorithms capable of achieving record-breaking results combined with embedded vision systems are the best candidate for the future of CV and video systems due to their flexibility and high accuracy in the area of image understanding. In this paper, we present an optimized convolutional network implementation suitable for real-time scene labeling on embedded platforms. We show that our algorithm can achieve up to 96GOp/s, running on the Nvidia Tegra K1 embedded SoC. We present experimental results, compare them to the state-of-the-art, and demonstrate that for scene labeling our approach achieves a 1.5x improvement in throughput when compared to a modern desktop CPU at a power budget of only 11 W.
Keywords :
computer vision; embedded systems; feedforward neural nets; image classification; CV; Nvidia Tegra K1; SoC; computer vision; convolutional network; desktop CPU; power 11 W; real-time embedded scene labeling; Convolution; Feature extraction; Graphics processing units; Labeling; Performance evaluation; Real-time systems; Throughput; Accelerator; Convolutional Networks; Scene Labeling;
Conference_Titel :
Design Automation Conference (DAC), 2015 52nd ACM/EDAC/IEEE
Conference_Location :
San Francisco, CA
DOI :
10.1145/2744769.2744788