Brain emulation circuit with reduced confusion

• US 4,773,024 A
• Filed: 06/03/1986
• Issued: 09/20/1988
• Est. Priority Date: 06/03/1986
• Status: Expired due to Term

• Current Assignee: MENO Corporation
• Original Assignee: Synaptics Incorporated
• Inventors: Faggin, Frederico, Sukonick, Josef S., Lynch, Gary S.
• Primary Examiner(s): Ruggiero, Joseph
• Application Number:
  US06/870,180
• Time in Patent Office: 840 Days
• Field of Search: 364/513, 364/148, 364/149-152, 364/156, 364/300, 364/200 MS File, 364/900 MS File, 382/14, 382/15, 382/27, 382/1, 382/36-39, 382/48, 382/50, 307/201
• US Class Current: 706/20
• CPC Class Codes: G06F 18/21 Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
• CPC Class Codes: G06N 3/063 using electronic means
• CPC Class Codes: G06V 10/75 Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries

Abstract

There is disclosed herein a recognize only embodiment of a recognition matrix comprised of a forward matrix and a reverse matrix each having a plurality of contacts which cause convergence responses on target lines when an input signal is received by said contact. Learning is performed by changing the characteristics of the contacts to alter the convergence responses they cause in accordance with a learning rule involving the comparison of total convergence response on each target line to a convergence threshold. The contacts are not programmed ad hoc in the field as events are individually learned. Instead each contact is programmed permanently by the user for a class of events which is fixed and which can never change. The user typically performs the learning on a computer simulator for all the events which a particular system is to be used to recognize. The patterns of convergence responses and contact structure characteristics which cause these convergence responses for the class of events as a whole are then examined and optimized for maximum recognition power and minimum confusion. This pattern of convergence responses or contact characteristics is then permanently programmed in the contacts of the forward and reverse matrices. A no-confusion embodiment is also disclosed whereby an array oif recognition machines are each programmed to recognize only one event, and all are coupled in parallel to an input bus carrying the signals characterizing the event to be recognized. The outputs are or'"'"'ed together.

Claims

1. A recognition matrix having a plurality of input lines and a forward matrix defined by said input lines and a plurality of target lines which terminate in convergence detecting circuits and a feedback matrix defined by said target lines and a plurality of feedback lines from the outputs of said convergence detecting circuits, each matrix having a plurality of contact structures coupled to selected target lines in said forward matrix, said recognition matrix characterized by the contact structures each being permanently programmed to have a characteristic to cause a predetermined convergence response on its associated target line in response to an active input signal received by said contact structure said predetermined characteristics being programmed by the user after optimization of the convergence response patterns of all said contacts for a given class of events to be learned.

2. A system for recognizing any of a plurality of input events comprising:

an input bus for carrying a plurality of input signals characterizing an event;

an array of recognition machines each having data inputs coupled to said input bus and each trained to recognize one event, each said recognition machine comprising;

an association matrix having a forward matrix and using a convergence rule to create an output vector and further comprising means for altering the convergence threshold used in implementing said convergence rule;

means in said means for altering the convergence threshold for automatically altering said convergence threshold on successive repetitions of said input event until recognition occurs or some other predetermined condition is met; and

characterized by the fact that the forward matrix of said association matrix is sparsely populated with randomly spaced contacts.