MAT: a tool for L2 pronunciation errors annotation
In the area of Computer Assisted Language Learning(CALL), second language (L2) learners spoken data is an important resource for analysing and annotating typical L2 pronunciation errors. The annotation of L2 pronunciation errors in spoken data is not an easy task though, normally it requires manual annotation from trained linguists or phoneticians. In order to facilitate this task, in this paper, we present the MAT tool, a web-based tool intended to facilitate the annotation of L2 learners{'} pronunciation errors at various levels. The tool has been designed taking into account recent studies on error detection in pronunciation training. It also aims at providing an easy and fast annotation process via a comprehensive and friendly user interface. The tool is based on the MARY TTS open source platform, from which it uses the components: text analyser (tokeniser, syllabifier, phonemiser), phonetic aligner and speech signal processor. Annotation results at sentence, word, syllable and phoneme levels are stored in XML format. The tool is currently under evaluation with a L2 learners spoken corpus recorded in the SPRINTER (Language Technology for Interactive, Multi-Media Online Language Learning) project.
Code (0)
등록된 구현이 없습니다.
Tasks
SentenceSimilar Papers 제목 키워드 기반
Automated detection of pronunciation errors in non-native English speech employing deep learning
Despite significant advances in recent years, the existing Computer-Assisted Pronunciation Training (CAPT) methods detect pronunciation errors with a relatively low accuracy (precision of 60% at 40%-80% recall). This Ph.…
Speech SynthesisTransducers with Pronunciation-aware Embeddings for Automatic Speech Recognition
This paper proposes Transducers with Pronunciation-aware Embeddings (PET). Unlike conventional Transducers where the decoder embeddings for different tokens are trained independently, the PET model's decoder embedding in…
Automatic Speech RecognitionDecoderspeech-recognitionSpeech RecognitionAUTOMATIC PRONUNCIATION MISTAKE DETECTOR PROJECT REPORT
Given the drawbacks of traditional English pronunciation correction systems, such as failure to provide timely feedback and correct learners' pronunciation errors, slow improvement of learners' English proficiency, and e…
Mistake Detectionspeech-recognitionSpeech RecognitionPhonological Level wav2vec2-based Mispronunciation Detection and Diagnosis Method
The automatic identification and analysis of pronunciation errors, known as Mispronunciation Detection and Diagnosis (MDD) plays a crucial role in Computer Aided Pronunciation Learning (CAPL) tools such as Second-Languag…
AttributeDiagnosticSonoEdit: Null-Space Constrained Knowledge Editing for Pronunciation Correction in LLM-Based TTS
Neural text-to-speech (TTS) systems systematically mispronounce low-resource proper nouns, particularly non-English names, brands, and geographic locations, due to their underrepresentation in predominantly English train…
knowledge editing