A Cmos Image Sensor And An Ai Accelerator For Realizing Edgecomputingbased Surveillance Camera Systems

• doi: 10.23919/VLSICircuits52068.2021.9492514
• title: A CMOS Image Sensor and an AI Accelerator for Realizing Edge-Computing-Based Surveillance Camera Systems
• publisher: IEEE
• isbn: 978-1-6654-4766-9
• issn: 2158-5601
• rank: 550
• access_type: LOCKED
• content_type: Conferences
• abstract: This paper presents a CMOS image sensor and an AI accelerator to realize surveillance camera systems based on edge computing. For CMOS image sensors to be used for surveillance, it is desirable that they are highly sensitive even in low illuminance. We propose a new timing shift ADC used in CMOS image sensors for improving high sensitivity performance. Our proposed ADC improves non-linearity characteristics under low illuminance by 63%. Achieving power-efficient edge computing is a challenge for the systems to be used widely in the surveillance camera market. We demonstrate that our proposed AI accelerator performs inference processing for object recognition with 1 TOPS/W. Keywords: CMOS image sensor, surveillance camera system, low light imaging, AI accelerator, edge computing
• article_number: 9492514
• pdf_url: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9492514
• html_url: https://ieeexplore.ieee.org/document/9492514/
• abstract_url: https://ieeexplore.ieee.org/document/9492514/
• publication_title: 2021 Symposium on VLSI Circuits
• conference_location: Kyoto, Japan
• conference_dates: 13-19 June 2021
• publication_number: 9492225
• is_number: 9492226
• publication_year: 2021
• publication_date: 13-19 June 2021
• start_page: 1
• end_page: 2
• citing_paper_count: 11
• citing_patent_count: 0
• download_count: 1386
• insert_date: 20210728
• index_terms:
  • ieee_terms:
    • Sensitivity
    • Surveillance
    • Image edge detection
    • AI accelerators
    • CMOS image sensors
    • Very large scale integration
    • Cameras
  • dynamic_index_terms:
    • Image Sensor
    • Camera Sensor
    • Surveillance Cameras
    • Surveillance Camera Systems
    • Low Light
    • Time Shift
    • Edge Computing
    • Edge Cloud
    • Low Illumination
    • Low-light Image
    • Deep Neural Network
    • Supply Voltage
• isbn_formats:
  • format: Print on Demand(PoD) ISBN, value: 978-1-6654-4766-9, isbnType: New-2005
  • format: Electronic ISBN, value: 978-4-86348-780-2, isbnType: New-2005
• authors:
  • Author Name: Fukashi Morishita
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37269135900
    ID: 37269135900
    Order: 1
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Norihito Kato
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37391260700
    ID: 37391260700
    Order: 2
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Satoshi Okubo
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37088921044
    ID: 37088921044
    Order: 3
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Takao Toi
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37265563300
    ID: 37265563300
    Order: 4
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Mitsuru Hiraki
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37354418400
    ID: 37354418400
    Order: 5
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Sugako Otani
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37320737900
    ID: 37320737900
    Order: 6
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Hideaki Abe
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37088919341
    ID: 37088919341
    Order: 7
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Yuji Shinohara
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37088920176
    ID: 37088920176
    Order: 8
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan
  • Author Name: Hiroyuki Kondo
    Affiliation: Renesas Electronics Corporation, Kodaira, Tokyo, Japan
    Author URL: https://ieeexplore.ieee.org/author/37320984200
    ID: 37320984200
    Order: 9
    Author Affiliations:
    • Renesas Electronics Corporation, Kodaira, Tokyo, Japan

Image Sensor

• sensor_type: CMOS
• resolution: 3840 x 2160 (4K)
• dynamic_range: Not specified
• pixel_size: 1.85μm x 1.85μm
• dark_current: Not specified

Optical Data

• focal_length: Not specified
• aperture: Not specified
• field_of_view: Not specified
• distortion: Not specified

Performance Metrics

• frame_rate: 60fps

Applications & Benefits

• cell_imaging: Not specified
• benefits: High sensitivity in low illuminance, wide dynamic range, power-efficient AI accelerators for object detection.

Supporting Organizations

• supported_by: Renesas Electronics Corporation, Kodaira, Tokyo, Japan

Manuscript Details

• publication_date: 13-19 June 2021

Relevancy Score

• score: 9
• missing_fields:
  • dark_current
  • dynamic_range
  • signal_to_noise_ratio
  • sensitivity
  • shutter_speed
  • power_consumption
  • readout_speed
  • noise_sources
  • focal_length
  • aperture
  • field_of_view
  • distortion
  • cell_imaging
  • benefits

Processed JSON Filename

request_2174a5bb-fd54-46ae-a829-42adafed0399-a_cmos_image_sensor_and_an_ai_accelerator_for_realizing_edgecomputingbased_surveillance_camera_systems.json