From Modern Standard Arabic to Levantine ASR: Leveraging GALE for dialects
Authors:
- 1 December 2011
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 266-271
Abstract
We report a series of experiments about how we can progress from Modern Standard Arabic (MSA) to Levantine ASR, in the context of the GALE DARPA program. While our GALE models achieved very low error rates, we still see error rates twice as high when decoding dialectal data. In this paper, we make use of a state-of-the-art Arabic dialect recognition system to automatically identify Levantine and MSA subsets in mixed speech of a variety of dialects including MSA. Training separate models on these subsets, we show a significant reduction in word error rate over using the entire data set to train one system for both dialects. During decoding, we use a tree array structure to mix Levantine and MSA models automatically using the posterior probabilities of the dialect classifier as soft weights. This technique allows us to mix these models without sacrificing performance for either varieties. Furthermore, using the initial acoustic-based dialect recognition system's output, we show that we can bootstrap a text-based dialect classifier and use it to identify relevant text data for building Levantine language models. Moreover, we compare different vowelization approaches when transitioning from MSA to Levantine models.Keywords
This publication has 12 references indexed in Scilit:
- The IBM 2011 GALE Arabic speech transcription system2011
- The IBM 2009 GALE Arabic speech transcription system2011
- The IBM Attila speech recognition toolkit2010
- Advances in Arabic Speech Transcription at IBM Under the DARPA GALE ProgramIEEE Transactions on Audio, Speech, and Language Processing, 2009
- Improving the Arabic pronunciation dictionary for phone and word recognition with linguistically-based pronunciation rules2009
- SVM Based Speaker Verification using a GMM Supervector Kernel and NAP Variability Compensation2006
- Support vector machines using GMM supervectors for speaker verificationIEEE Signal Processing Letters, 2006
- Arabic tokenization, part-of-speech tagging and morphological disambiguation in one fell swoop2005
- Cross-dialectal acoustic data sharing for Arabic speech recognition2004
- Speaker Verification Using Adapted Gaussian Mixture ModelsDigital Signal Processing, 2000