The Community for Technology Leaders
RSS Icon
Subscribe
Issue No.06 - Nov.-Dec. (2012 vol.32)
pp: 38-50
Jinwook Oh , Korea Advanced Institute of Science and Technology
Gyeonghoon Kim , Korea Advanced Institute of Science and Technology
Injoon Hong , Korea Advanced Institute of Science and Technology
Junyoung Park , Korea Advanced Institute of Science and Technology
Seungjin Lee , Korea Advanced Institute of Science and Technology
Joo-Young Kim , Korea Advanced Institute of Science and Technology
Jeong-Ho Woo , Korea Advanced Institute of Science and Technology
Hoi-Jun Yoo , Korea Advanced Institute of Science and Technology
ABSTRACT
A new low-power object-recognition processor achieves real-time robust recognition, satisfying modern mobile vision systems' requirements. The authors introduce an attention-based object-recognition algorithm for energy efficiency, a heterogeneous multicore architecture for data- and thread-level parallelism, and a network on a chip for high on-chip bandwidth. The fabricated chip achieves 30 frames/second throughput and an average 320 mW power consumption on test 720p video sequences, yielding 640 GOPS/W and 10.5 nJ/pixel energy efficiency.
INDEX TERMS
Decision support systems, Robustness, Object recognition, Multicore processing, Network-on-a-chip, Low power electronics, SIFT, object recognition, attention, attention-based object recognition, network-on-chip, multicore processor, heterogeneous multicore, object-recognition pipeline, scale invariant feature transform
CITATION
Jinwook Oh, Gyeonghoon Kim, Injoon Hong, Junyoung Park, Seungjin Lee, Joo-Young Kim, Jeong-Ho Woo, Hoi-Jun Yoo, "Low-Power, Real-Time Object-Recognition Processors for Mobile Vision Systems", IEEE Micro, vol.32, no. 6, pp. 38-50, Nov.-Dec. 2012, doi:10.1109/MM.2012.90
REFERENCES
1. D.G. Lowe, "Distinctive Image Features from Scale-Invariant Keypoints," ACM Int'l J. Computer Vision, Jan. 2004, pp. 91-110.
2. S.N. Sinha et al., "GPU-Based Video Feature Tracking and Matching," Workshop on Edge Computing Using New Commodity Architectures, 2006.
3. J.-Y. Kim et al., "A 201.4GOPS 496mW Real-Time Multi-Object Recognition Processor with Bio-Inspired Neural Perception Engine," Proc. IEEE Int'l Solid-State Circuits Conf., IEEE CS, 2008, pp. 150-151.
4. S. Lee et al., "A 345mW Heterogeneous Many-Core Processor with an Intelligent Inference Engine for Robust Object Recognition," Proc. IEEE J. Solid-State Circuits, Jan. 2011, pp. 42-51.
5. J. Oh et al., "A 320mW 342GOPS Real-Time Moving Object Recognition Processor for HD 720p Video Streams," Proc. IEEE Int'l Solid-State Circuits Conf., IEEE CS, 2012, pp. 220-221.
6. J. Clemons et al., "MEVBench: A Mobile Computer Vision Benchmarking Suite," Proc. IEEE Int'l Conf. Workload Characterization, IEEE CS, 2011, pp. 91-102.
7. Y. Tanabe et al., "A 464GOPS 620GOPS/W Heterogeneous Multi-Core SoC for Image-Recognition Applications," IEEE Int'l Solid-State Circuits Conf., IEEE CS, 2012, pp. 222-223.
8. J. Clemons et al., "EFFEX: An Embedded Processor for Computer Vision Based Feature Extraction," Proc. IEEE Design Automation Conf., IEEE CS, 2011, pp. 1020-1025.
33 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool