N-Shot Benchmarking of Whisper on Diverse Arabic Speech Recognition

• URL: http://arxiv.org/abs/2306.02902v1
• Date: Mon, 5 Jun 2023 14:09:25 GMT
• Title: N-Shot Benchmarking of Whisper on Diverse Arabic Speech Recognition
• Authors: Bashar Talafha, Abdul Waheed, Muhammad Abdul-Mageed
• Abstract summary: It is not clear how Whisper would fare under diverse conditions even on languages it was evaluated on such as Arabic. Our evaluation covers most publicly available Arabic speech data and is performed under n-shot finetuning. We also investigate the robustness of Whisper under completely novel conditions, such as in dialect-accented standard Arabic and in unseen dialects.
• Score: 8.731646409966737
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Whisper, the recently developed multilingual weakly supervised model, is reported to perform well on multiple speech recognition benchmarks in both monolingual and multilingual settings. However, it is not clear how Whisper would fare under diverse conditions even on languages it was evaluated on such as Arabic. In this work, we address this gap by comprehensively evaluating Whisper on several varieties of Arabic speech for the ASR task. Our evaluation covers most publicly available Arabic speech data and is performed under n-shot (zero-, few-, and full) finetuning. We also investigate the robustness of Whisper under completely novel conditions, such as in dialect-accented standard Arabic and in unseen dialects for which we develop evaluation data. Our experiments show that although Whisper zero-shot outperforms fully finetuned XLS-R models on all datasets, its performance deteriorates significantly in the zero-shot setting for five unseen dialects (i.e., Algeria, Jordan, Palestine, UAE, and Yemen).

Related papers

• Cross-Lingual Transfer Learning for Speech Translation [7.802021866251242]
  Zero-shot cross-lingual transfer has been demonstrated on a range of NLP tasks. We explore whether speech-based models exhibit the same transfer capability.
  arXiv  Detail & Related papers  (2024-07-01T09:51:48Z)
• Voices Unheard: NLP Resources and Models for Yorùbá Regional Dialects [72.18753241750964]
  Yorub'a is an African language with roughly 47 million speakers. Recent efforts to develop NLP technologies for African languages have focused on their standard dialects. We take steps towards bridging this gap by introducing a new high-quality parallel text and speech corpus.
  arXiv  Detail & Related papers  (2024-06-27T22:38:04Z)
• An Initial Investigation of Language Adaptation for TTS Systems under Low-resource Scenarios [76.11409260727459]
  This paper explores the language adaptation capability of ZMM-TTS, a recent SSL-based multilingual TTS system. We demonstrate that the similarity in phonetics between the pre-training and target languages, as well as the language category, affects the target language's adaptation performance.
  arXiv  Detail & Related papers  (2024-06-13T08:16:52Z)
• ArabicMMLU: Assessing Massive Multitask Language Understanding in Arabic [53.1913348687902]
  We present ArabicMMLU, the first multi-task language understanding benchmark for Arabic language. Our data comprises 40 tasks and 14,575 multiple-choice questions in Modern Standard Arabic (MSA) Our evaluations of 35 models reveal substantial room for improvement, particularly among the best open-source models.
  arXiv  Detail & Related papers  (2024-02-20T09:07:41Z)
• ALDi: Quantifying the Arabic Level of Dialectness of Text [17.37857915257019]
  We argue that Arabic speakers perceive a spectrum of dialectness, which we operationalize at the sentence level as the Arabic Level of Dialectness (ALDi) We provide a detailed analysis of AOC-ALDi and show that a model trained on it can effectively identify levels of dialectness on a range of other corpora.
  arXiv  Detail & Related papers  (2023-10-20T18:07:39Z)
• VoxArabica: A Robust Dialect-Aware Arabic Speech Recognition System [16.420831300734697]
  VoxArabica is a system for dialect identification (DID) and automatic speech recognition (ASR) of Arabic. We train a wide range of models such as HuBERT (DID), Whisper, and XLS-R (ASR) in a supervised setting for Arabic DID and ASR tasks. We finetune our ASR models on MSA, Egyptian, Moroccan, and mixed data. We integrate these models into a single web interface with diverse features such as audio recording, file upload, model selection, and the option to raise flags for incorrect outputs.
  arXiv  Detail & Related papers  (2023-10-17T08:33:02Z)
• NusaWrites: Constructing High-Quality Corpora for Underrepresented and Extremely Low-Resource Languages [54.808217147579036]
  We conduct a case study on Indonesian local languages. We compare the effectiveness of online scraping, human translation, and paragraph writing by native speakers in constructing datasets. Our findings demonstrate that datasets generated through paragraph writing by native speakers exhibit superior quality in terms of lexical diversity and cultural content.
  arXiv  Detail & Related papers  (2023-09-19T14:42:33Z)
• Offensive Language Detection in Under-resourced Algerian Dialectal Arabic Language [0.0]
  We focus on the Algerian dialectal Arabic which is one of under-resourced languages. Due to the scarcity of works on the same language, we have built a new corpus regrouping more than 8.7k texts manually annotated as normal, abusive and offensive.
  arXiv  Detail & Related papers  (2022-03-18T15:42:21Z)
• Towards One Model to Rule All: Multilingual Strategy for Dialectal Code-Switching Arabic ASR [11.363966269198064]
  We design a large multilingual end-to-end ASR using self-attention based conformer architecture. We trained the system using Arabic (Ar), English (En) and French (Fr) languages. Our findings demonstrate the strength of such a model by outperforming state-of-the-art monolingual dialectal Arabic and code-switching Arabic ASR.
  arXiv  Detail & Related papers  (2021-05-31T08:20:38Z)
• How Phonotactics Affect Multilingual and Zero-shot ASR Performance [74.70048598292583]
  A Transformer encoder-decoder model has been shown to leverage multilingual data well in IPA transcriptions of languages presented during training. We replace the encoder-decoder with a hybrid ASR system consisting of a separate AM and LM. We show that the gain from modeling crosslingual phonotactics is limited, and imposing a too strong model can hurt the zero-shot transfer.
  arXiv  Detail & Related papers  (2020-10-22T23:07:24Z)
• That Sounds Familiar: an Analysis of Phonetic Representations Transfer Across Languages [72.9927937955371]
  We use the resources existing in other languages to train a multilingual automatic speech recognition model. We observe significant improvements across all languages in the multilingual setting, and stark degradation in the crosslingual setting. Our analysis uncovered that even the phones that are unique to a single language can benefit greatly from adding training data from other languages.
  arXiv  Detail & Related papers  (2020-05-16T22:28:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.