NUTS, NARS, and Speech

TL;DR

This paper introduces NUTS, a novel non-axiomatic reasoning approach for speech recognition that achieves competitive performance with minimal training data, demonstrating adaptability and efficiency.

Contribution

It presents NUTS, combining naive dimensionality reduction and NARS, as a new few-shot speech recognition method that operates effectively with only two training examples.

Findings

NUTS performs similarly to Whisper Tiny with just 2 training examples.

NUTS utilizes non-axiomatic reasoning for speech perception.

The approach demonstrates high adaptability with limited data.

Abstract

To investigate whether "Intelligence is the capacity of an information-processing system to adapt to its environment while operating with insufficient knowledge and resources", we look at utilising the non axiomatic reasoning system (NARS) for speech recognition. This article presents NUTS: raNdom dimensionality redUction non axiomaTic reasoning few Shot learner for perception. NUTS consists of naive dimensionality reduction, some pre-processing, and then non axiomatic reasoning (NARS). With only 2 training examples NUTS performs similarly to the Whisper Tiny model for discrete word identification.