Sub-Microwatt Analog VLSI Trainable Pattern Classifier

Abstract

The design and implementation of an analog system-on-chip template-based pattern classifier for biometric signature verification at sub-microwatt power is presented. A pro- grammable array of floating-gate subthreshold MOS translinear circuits matches input features with stored templates and com- bines the scores into category outputs. Subtractive normalization of the outputs by current-mode feedback produces confidence scores which are integrated for category selection. The classifier implements a support vector machine to select programming values from training samples. A two-step calibration procedure during programming alleviates offset and gain errors in the analog array. A 24-class, 14-input, 720-template classifier trained for speaker identification and fabricated on a 3 mm 3 mm chip in 0.5 m CMOS delivers real-time recognition accuracy on par with floating-point emulation in software. At 40 classifications per second and 840 nW power, the processor attains a computational efficiency of /49 /51 /49/48/49/50multiply-accumulates per second per Watt of power.