Errgrams - A Way to Improving ASR for Highly Inflected Dravidian Languages
نویسندگان
چکیده
In this paper, we present results of our experiments with ASR for a highly inflected Dravidian language, Telugu. First, we propose a new metric for evaluating ASR performance for inflectional languages (Inflectional Word Error Rate IWER) which takes into account whether the incorrectly recognized word corresponds to the same lexicon lemma or not. We also present results achieved by applying a novel method – errgrams – to ASR lattice. With respect to confidence scores, the method tries to learn typical error patterns, which are then used for lattice correction, and applied just before standard lattice rescoring. Our confidence measures are based on word posteriors and were improved by applying antimodels trained on anti-examples generated by the standard N-gram language model. For Telugu language, we decreased the WER from 45.2% to 40.4% (by 4.8% absolute), and the IWER from 41.6% to 39.5% (2.1 % absolute), with respect to the baseline performance. All improvements are statistically significant using all three standard NIST significance tests for ASR.
منابع مشابه
ASR for South Slavic Languages Developed in Almost Automated Way
Slavic languages pose several specific challenges that need to be addressed in an ASR system design. Since we have already built an engine suited for highly-inflected languages, we focus on adopting it for new languages, now. In this case, we present an efficient way to adapt the system to all (seven) South Slavic languages, using methods and tools that benefit from language similarities, easil...
متن کاملSpecifications of Building Polish Lexica for Application in ASR and TTS Systems
This paper brings detailed information concerning the specifications of building Polish lexica of common and special application words for use in speech applications such as ASR (automatic speech recognition) or TTS (text-to-speech) synthesis. The specifications include information on the collection of text corpora and word lists, phonetic, grammatical and morphological annotation, as well as s...
متن کاملSignificance of an Accurate Sandhi-Splitter in Shallow Parsing of Dravidian Languages
This paper evaluates the challenges involved in shallow parsing of Dravidian languages which are highly agglutinative and morphologically rich. Text processing tasks in these languages are not trivial because multiple words concatenate to form a single string with morpho-phonemic changes at the point of concatenation. This phenomenon known as Sandhi, in turn complicates the individual word iden...
متن کاملHandling verb phrase morphology in highly inflected Indian languages for Machine Translation
The phrase based systems for machine translation are limited by the phrases that they see during the training. For highly inflected languages, it is uncommon to see all the forms of a word in the parallel corpora used during training. This problem is amplified for verbs in highly inflected languages where the correct form of the word depends on factors like gender, number and tense aspect. We p...
متن کاملMorphological Analyzer for Affix Stacking Languages: A Case Study of Marathi
In this paper we describe and evaluate a Finite State Machine (FSM) based Morphological Analyzer (MA) for Marathi, a highly inflectional language with agglutinative suffixes. Marathi belongs to the Indo-European family and is considerably influenced by Dravidian languages. Adroit handling of participial constructions and other derived forms (Krudantas and Taddhitas) in addition to inflected for...
متن کامل