Conference Paper

A Two-Dimensional, Object-Based Analog VLSI Visual Attention System.

Sch. of Electr. & Comput. Eng., Georgia Inst. of Technol., Atlanta, GA;
DOI: 10.1109/ARVLSI.1999.756055 Conference: 18th Conference on Advanced Research in VLSI (ARVLSI '99), 21-24 March 1999, Atlanta, GA, USA
Source: DBLP

ABSTRACT A two-dimensional object-based analog VLSI model of selective attentional processing has been implemented using a standard 1.2 μm CMOS process. This chip extends previous work modeling object-based selection and scanning by incorporating the circuity and architectural changes necessary for two-dimensional focal plane processing. To balance the need for closely spaced large photodetectors with the space requirements of complex in-pixel processing, the chip implements a multiresolution architecture. The system has he ability to group pixels into objects; this grouping is dynamic, driven solely by the segmentation criterion at the input. In the demonstration system, image intensity has been chosen for the input saliency map and the segmentation is based on spatial lowpass filtering followed by an intensity threshold. We present experimental results

1 Bookmark
  • [Show abstract] [Hide abstract]
    ABSTRACT: A biologically-inspired hybrid vision chip is presented for real-time object-based processing for tasks such as centroiding, sizing and counting of enclosed objects. This system presents the first silicon retina capable of centroiding and sizing multiple objects in true parallel fashion. Based on a novel distributed algorithm, this approach uses the input image to enclose a feedback loop to realize a data-driven pulsating action. The sensor provides a resolution of 48 × 48 pixels with a 85 μm×85 μm pixel footprint and has been measured to consume 243 μW at 1.8-V supply, achieving an equivalent computational efficiency of 724.64 MIPS/mW with a 500-μs process time.
    IEEE Journal of Solid-State Circuits 07/2006; · 3.06 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A CMOS high-performance current-mode winner-take-all circuit is presented. The circuit employs a novel technique for inhibitory and excitatory feedbacks based on input currents average computation, achieving both high speed and high precision. The circuit is designed for operation with a wide range of input current values, allowing its integration with circuits operating both in subthreshold and in strong inversion regions. Two circuits, each for a different range of input currents, have been implemented in a standard 0.35-μm CMOS process available through MOSIS and are operated via a 3.3-V supply. Their operation is discussed, simulation results are reported and preliminary measurements from a test chip are presented.
    Circuits and Systems II: Express Briefs, IEEE Transactions on 04/2005; · 1.33 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: An architecture for implementation of a novel tracking VLSI sensor for multiple target detection and tracking is presented. The sensor, based on the proposed implementation concept, allows acquisition and real time tracking of up to three bright targets in the field of view. While based on the spotlight model of visual attention in biological systems, the proposed sensor features several advantages. This includes distractors elimination, attentional shifts possibility with no dependence on the distance between the targets of interest, high quality image in the snapshot mode of operation simultaneously with tracking and low-power dissipation. A comparison of the proposed concept to the existing spotlight and object-based visual attention models is discussed and a brief description of the proposed sensor is given.
    Electronics, Circuits and Systems, 2004. ICECS 2004. Proceedings of the 2004 11th IEEE International Conference on; 01/2005

Full-text (2 Sources)

Available from
Jun 1, 2014