MACSen: A Processing-In-Sensor Architecture Integrating MAC Operations Into Image Sensor for Ultra-Low-Power BNN-Based Intelligent Visual Perception
In: IEEE Transactions on Circuits and Systems II: Express Briefs, Jg. 68 (2021-02-01), S. 627-631
Online
unknown
Zugriff:
Current BNN-based visual system wastes lots of energy in data conversion and movement, hindering its deployment on battery-powered devices. This brief proposes MACSen, an ultra-low-power processing-in-sensor (PIS) architecture which integrates sensing with computing and directly outputs the computation results. The multiply-and-accumulation (MAC) operation in BNN is fused with the Correlated Double Sampling (CDS) procedure together to save data conversion power. A $4\times 4$ MACSen prototype of 180nm process was fabricated for demonstration, and it achieves the frame rate of 1000fps and the energy efficiency of 1.32TOP/s/W in computation mode. Furthermore, the system demonstration on MNIST dataset classification task shows that the hardware BNN implementation integrating MACSen incurs no accuracy degradation and gains 61% energy saving compared with state-of-the-art work.
Titel: |
MACSen: A Processing-In-Sensor Architecture Integrating MAC Operations Into Image Sensor for Ultra-Low-Power BNN-Based Intelligent Visual Perception
|
---|---|
Autor/in / Beteiligte Person: | Lin, Ningchao ; Wei, Qi ; Li, Ziru ; Yang, Huazhong ; Xu, Han ; Qiao, Fei ; Yin, Xunzhao |
Link: | |
Zeitschrift: | IEEE Transactions on Circuits and Systems II: Express Briefs, Jg. 68 (2021-02-01), S. 627-631 |
Veröffentlichung: | Institute of Electrical and Electronics Engineers (IEEE), 2021 |
Medientyp: | unknown |
ISSN: | 1558-3791 (print) ; 1549-7747 (print) |
DOI: | 10.1109/tcsii.2020.3015902 |
Schlagwort: |
|
Sonstiges: |
|