Abstract
This paper presents an always-on Complementary Metal Oxide Semiconductor (CMOS) image sensor (CIS) using an analog convolutional neural network for image classification in mobile applications. To reduce the power consumption as well as the overall processing time, we propose analog convolution circuits for computing convolution, max-pooling, and correlated double sampling operations without operational transconductance amplifiers. In addition, we used the voltage-mode MAX circuit for max pooling in the analog domain. After the analog convolution processing, the image data were reduced by 99.58% and were converted to digital with a 4-bit single-slope analog-to-digital converter. After the conversion, images were classified by the fully connected processor, which is traditionally performed in the digital domain. The measurement results show that we achieved an 89.33% image classification accuracy. The prototype CIS was fabricated in a 0.11 μm 1-poly 4-metal CIS process with a standard 4T-active pixel sensor. The image resolution was 160 × 120, and the total power consumption of the proposed CIS was 1.12 mW with a 3.3 V supply voltage and a maximum frame rate of 120.
Highlights
In recent years, as the number of smart devices has increased with the rise of the Internet of Things [1], the importance of user authentication has increased as well
We propose an always-on Complementary Metal Oxide Semiconductor image sensor (CIS) based on an analog lightweight convolutional neural network (LWCNN) for image classification
Using the proposed CIS, images can be classified without a high-resolution analog-to-digital converters (ADC) or additional memory blocks for convolutional neural network (CNN) processing
Summary
As the number of smart devices has increased with the rise of the Internet of Things [1], the importance of user authentication has increased as well. As an example of user-authentication applications, always-on face detection/recognition is highly convenient because direct physical contact, such as fingerprint scanning, is unnecessary [2,3,4]. Integrating always-on face detection/recognition into mobile devices is challenging because of these devices’. Complementary Metal Oxide Semiconductor (CMOS) image sensor (CIS) that enables high power consuming devices, like ultra-high resolution (>tens of megapixels) CISs, to turn on for iris identification and face identification have received great attention [5]. In the CIS, the light intensity that accumulates in the pixel array is converted into the corresponding voltage, which is transmitted into the digital domain with column-parallel analog-to-digital converters (ADC). The pixel data are transferred to an external CVP chip and stored in analog memory blocks before a complex deep convolutional neural network (CNN) operation in the CVP that allows the Sensors 2020, 20, 3101; doi:10.3390/s20113101 www.mdpi.com/journal/sensors
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
