Bu çalışmada konuşmadan metne çeviri için önerilmiş ve çok sayıda dille ön eğitilmiş iki model olan Whisper-Small ve Wav2Vec2-XLS-R-300M modellerinin Türkçe dilinde konuşmadan metne çevirme başarıları incelenmiştir. Çalışmada açık kaynaklı bir veri kümesi olan Türkçe dilinde hazırlanmış Mozilla Common Voice 11.0 versiyonu kullanılmıştır. Az sayıda veri içeren bu veri kümesi ile çok dilli modeller olan Whisper-Small ve Wav2Vec2-XLS-R-300M ince ayar yapılmıştır. İki modelin konuşmadan metne çeviri başarımı değerlendirilmiş ve Wav2Vec2-XLS-R-300M modelinin 0,28 WER değeri Whisper-Small modelinin 0,16 WER değeri gösterdiği gözlemlenmiştir. Ek olarak modellerin başarısı eğitim ve doğrulama veri kümesinde bulunmayan çağrı merkezi kayıtlarıyla hazırlanmış sınama verisiyle incelenmiştir.
TÜBİTAK TEYDEB 1501
3210713
Bu çalışma TÜBİTAK TEYDEB 1501 kapsamında desteklenmekte olan 3210713 numaralı "Güncel Derin Öğrenme Mimarileri ile Türkçe Dili için Konuşmadan Metne Çeviri Yapabilen ve Hizmet Olarak Yazılım (SaaS) Modeli ile Çalışan Sistemin Geliştirilmesi" isimli proje kapsamında gerçekleştirilmiştir.
In this study, the performances of the Whisper-Small and Wav2Vec2-XLS-R-300M models which are two pre-trained multilingual models for speech to text were examined for the Turkish language. Mozilla Common Voice version 11.0 which is prepared in Turkish language and is an open-source data set, was used in the study. The multilingual models, Whisper-Small and Wav2Vec2-XLS-R-300M were fine-tuned with this data set which contains a small amount of data. The speech to text performance of the two models was compared. WER values are calculated as 0.28 and 0.16 for the Wav2Vec2-XLS-R-300M and the Whisper-Small models respectively. In addition, the performances of the models were examined with the test data prepared with call center records that were not included in the training and validation dataset.
3210713
Primary Language | Turkish |
---|---|
Subjects | Engineering |
Journal Section | Makaleler(Araştırma) |
Authors | |
Project Number | 3210713 |
Early Pub Date | October 22, 2023 |
Publication Date | November 20, 2023 |
Published in Issue | Year 2023 Volume: 16 Issue: 2 |
Article Acceptance
Use user registration/login to upload articles online.
The acceptance process of the articles sent to the journal consists of the following stages:
1. Each submitted article is sent to at least two referees at the first stage.
2. Referee appointments are made by the journal editors. There are approximately 200 referees in the referee pool of the journal and these referees are classified according to their areas of interest. Each referee is sent an article on the subject he is interested in. The selection of the arbitrator is done in a way that does not cause any conflict of interest.
3. In the articles sent to the referees, the names of the authors are closed.
4. Referees are explained how to evaluate an article and are asked to fill in the evaluation form shown below.
5. The articles in which two referees give positive opinion are subjected to similarity review by the editors. The similarity in the articles is expected to be less than 25%.
6. A paper that has passed all stages is reviewed by the editor in terms of language and presentation, and necessary corrections and improvements are made. If necessary, the authors are notified of the situation.
. This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.