Low-Power, Real-Time Object-Recognition Processors for Mobile Vision Systems

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-lev...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	IEEE MICRO 2012-11, Vol.32 (6), p.38-50
Hauptverfasser:	Oh, Jinwook, Kim, Gyeonghoon, Hong, Injoon, Park, Junyoung, Lee, Seungjin, Kim, Joo-Young, Woo, Jeong-Ho, Yoo, Hoi-Jun
Format:	Artikel
Sprache:	eng
Schlagworte:	attention attention-based object recognition Chip formation Chips Computer architecture Decision support systems Energy efficiency Energy management heterogeneous multicore Low power electronics Multicore processing multicore processor Network-on-a-chip network-on-chip Object recognition object-recognition pipeline Power consumption Processors Real time Robustness scale invariant feature transform SIFT Vision systems
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	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.
ISSN:	0272-1732 1937-4143
DOI:	10.1109/MM.2012.90