Language identification using a combined articulatory prosody framework

This study presents new advancements in our articulatory-based language identi cation (LID) system. Our LID system automatically identi es language-features (LFs) from a phonological features (PFs) based representation of speech. While our baseline system uses a static PF-representation for extracting LFs, the new system is based on a dynamic PF representation for feature extraction. Interestingly, the new LFs outperform our baseline system

• Type: Journal
• Year: 2011
• Where: ICASSP
• Authors: Abhijeet Sangwan, Mahnoosh Mehrabani, John H. L. Hansen
• Keywords: Dynamic Pf Representation, ICASSP 2011, Language Identi Cation, Phonological Features, Signal Processing