Acoustic Modeling Using a Shallow CNN-HTSVM Architecture

Autor(es)

Shulby, Christopher Dane ; Ferreira, Martha Dais ; de Mello, Rodrigo F. ; Aluisio, Sandra Maria ; IEEE

• Tipo de documento: Artigo Científico
• Fonte: 2017 6TH BRAZILIAN CONFERENCE ON INTELLIGENT SYSTEMS (BRACIS); v. N/A, p. 6-pg., 2017-01-01.

Resumo

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 (Gaussian Mixture Model -Hidden Markov Model) 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. (AU)

• Processo FAPESP: 13/07375-0 - CeMEAI - Centro de Ciências Matemáticas Aplicadas à Indústria
• Beneficiário: Francisco Louzada Neto
• Modalidade de apoio: Auxílio à Pesquisa - Centros de Pesquisa, Inovação e Difusão - CEPIDs