A VLSI neuroprocessor for image restoration using analog computing-based systolic architecture

An analog computing-based systolic architecture which employs multiple neuroprocessors for high-speed early vision processing is presented. For a two-dimensional image, parallel processing is performed in the row direction and pipelined processing is performed in the column direction. The mixed analog/digital design approach is suitable for implementation of electronic neural systems. Local data computation is executed by analog circuitry to achieve full parallelism and to minimize power dissipation. Inter-processor communication is carried out in the digital format to maintain strong signal strength across the chip boundary and to achieve direct scalability in neural network size. For demonstration purposes, a compact and efficient VLSI neural chip that includes multiple neuroprocessors for high-speed digital image restoration is designed. Measured results of the programmable synapse, and statistical distribution of measured synapse conductances are presented. Based on these results, system-level analyses at 8-bit resolution are conducted. A 8.0×6.0-mm2 chip from a 1.2-µm CMOS technology can accommodate 5 neuroprocessors and the speed-up factor over the Sun-4/75 SPARC workstation is around 450. This chip achieves 18 Giga connections per second.

[1]  B. K. Jenkins,et al.  Image restoration using a neural network , 1988, IEEE Trans. Acoust. Speech Signal Process..

[2]  S. Sakiyama,et al.  A 200 MIPS image signal multiprocessor on a single chip , 1990, 1990 37th IEEE International Conference on Solid-State Circuits.

[3]  Bing J. Sheu,et al.  An analog neural network processor for self-organizing mapping , 1992, 1992 IEEE International Solid-State Circuits Conference Digest of Technical Papers.

[4]  Paul Jespers,et al.  A class of multiprocessors for real-time image and multidimensional signal processing , 1988 .

[5]  John J. Hopfield,et al.  Simple 'neural' optimization networks: An A/D converter, signal decision circuit, and a linear programming circuit , 1986 .

[6]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[7]  Thomas S. Huang,et al.  Image processing , 1971 .

[8]  H. T. Kung.  Why systolic architectures? , 1982, Computer.

[9]  C Koch,et al.  Analog "neuronal" networks in early vision. , 1986, Proceedings of the National Academy of Sciences of the United States of America.

[10]  Mohammed Ismail,et al.  Issues in Analog VLSI and MOS Techniques for Neural Computing , 1989, Analog VLSI Implementation of Neural Systems.

[11]  C. Tomovich,et al.  MOSIS - A gateway to silicon , 1988, IEEE Circuits and Devices Magazine.

[12]  Tomaso Poggio,et al.  Computational vision and regularization theory , 1985, Nature.

[13]  Robert F. Miracky,et al.  Technologies for rapid prototyping of multi-chip modules , 1991, [1991 Proceedings] IEEE International Conference on Computer Design: VLSI in Computers and Processors.

[14]  R. Chellappa,et al.  Image restoration with neural networks , 1992 .

[15]  Neil Weste,et al.  Principles of CMOS VLSI Design , 1985 .

[16]  S. Tam,et al.  An electrically trainable artificial neural network (ETANN) with 10240 'floating gate' synapses , 1990, International 1989 Joint Conference on Neural Networks.

[17]  Mohamed El-Sharkawy,et al.  Real Time Digital Signal Processing Applications With Motorola's Dsp56000 Family , 1990 .

[18]  Bing J. Sheu,et al.  Design of a neural-based A/D converter using modified Hopfield network , 1989 .

[19]  B. R. Hunt,et al.  Digital Image Restoration , 1977 .

[20]  J. Goodman,et al.  Neural networks for computation: number representations and programming complexity. , 1986, Applied optics.

[21]  Bing J. Sheu,et al.  A nonvolatile analog neural memory using floating-gate MOS transistors , 1992 .

[22]  J. J. Paulos,et al.  Artificial neural networks using MOS analog multipliers , 1990 .

[23]  J. Woods,et al.  Kalman filtering in two dimensions: Further results , 1981 .

[24]  Richard P. Lippmann,et al.  An introduction to computing with neural nets , 1987 .

[25]  Bing J. Sheu,et al.  A compact and general-purpose neural chip with electrically programmable synapses , 1990, IEEE Proceedings of the Custom Integrated Circuits Conference.

[26]  J. Cosgrove,et al.  Array processors , 1980, IEEE Acoustics, Speech, and Signal Processing Newsletter.

[27]  Anil K. Jain.  Fundamentals of Digital Image Processing , 2018, Control of Color Imaging Systems.

[28]  T. Poggio,et al.  III-Posed problems early vision: from computational theory to analogue networks , 1985, Proceedings of the Royal Society of London. Series B. Biological Sciences.

[29]  Y. Tamura,et al.  A BiCMOS analog neural network with dynamically updated weights , 1992, 1990 37th IEEE International Conference on Solid-State Circuits.

[30]  JOHN w. WOODS,et al.  Kalman filtering in two dimensions , 1977, IEEE Trans. Inf. Theory.

[31]  Paul Wintz,et al.  Digital image processing (2nd ed.) , 1987 .

[32]  C. D. Gelatt,et al.  Optimization by Simulated Annealing , 1983, Science.

[33]  Bing J. Sheu,et al.  Hardware annealing in analog VLSI neurocomputing , 1990 .

[34]  Kai Hwang,et al.  Computer architecture and parallel processing , 1984, McGraw-Hill Series in computer organization and architecture.

[35]  S. Kung,et al.  VLSI Array processors , 1985, IEEE ASSP Magazine.