A Trainable Sequence Learner that Learns and Recognizes Two-Input Sequence Patterns

TL;DR

This paper introduces two analog circuit designs capable of learning and recognizing two-input temporal sequences, with different training and timing flexibility features, suitable for fast nanosecond-scale applications.

Contribution

It proposes novel analog circuit architectures for sequence learning and recognition, with one design supporting runtime reset and the other offering single training with flexible timing.

Findings

First design can reset training during runtime

Second design is trained once with flexible input timing

Both designs recognize sequences in tens of nanoseconds

Abstract

We present two designs for an analog circuit that can learn to detect a temporal sequence of two inputs. The training phase is done by feeding the circuit with the desired sequence and, after the training is completed, each time the trained sequence is encountered again the circuit will emit a signal of correct recognition. Sequences are in the order of tens of nanoseconds. The first design can reset the trained sequence on runtime but assumes very strict timing of the inputs. The second design can only be trained once but is lenient in the input's timing.