# Acoustic Modeling Using a Shallow CNN-HTSVM Architecture

## TL;DR

This paper introduces a shallow CNN-HTSVM architecture for speech recognition that leverages knowledge-based rules and machine learning, achieving high accuracy with limited data and outperforming traditional models.

## Contribution

It presents a novel shallow CNN-HTSVM hierarchical architecture that combines rule-based knowledge with statistical learning for improved acoustic modeling.

## Key findings

- Outperforms traditional GMM-HMM models
- HTSVM outperforms MLP classifiers
- Effective with small datasets

## Abstract

High-accuracy speech recognition is especially challenging when large datasets are not available. It is possible to bridge this gap with careful and knowledge-driven parsing combined with the biologically inspired CNN and the learning guarantees of the Vapnik Chervonenkis (VC) theory. This work presents a Shallow-CNN-HTSVM (Hierarchical Tree Support Vector Machine classifier) architecture which uses a predefined knowledge-based set of rules with statistical machine learning techniques. Here we show that gross errors present even in state-of-the-art systems can be avoided and that an accurate acoustic model can be built in a hierarchical fashion. The CNN-HTSVM acoustic model outperforms traditional GMM-HMM models and the HTSVM structure outperforms a MLP multi-class classifier. More importantly we isolate the performance of the acoustic model and provide results on both the frame and phoneme level considering the true robustness of the model. We show that even with a small amount of data accurate and robust recognition rates can be obtained.