Dynamic acoustic unit augmentation with bpe-dropout for low-resource end-to-end speech recognition
With the rapid development of speech assistants, adapting server-intended automatic
speech recognition (ASR) solutions to a direct device has become crucial. For on-device …
speech recognition (ASR) solutions to a direct device has become crucial. For on-device …
The University of Helsinki and Aalto University submissions to the WMT 2020 news and low-resource translation tasks
This paper describes the joint participation of University of Helsinki and Aalto University to
two shared tasks of WMT 2020: the news translation between Inuktitut and English and the …
two shared tasks of WMT 2020: the news translation between Inuktitut and English and the …
Boosting neural machine translation from finnish to northern Sámi with rule-based backtranslation
We consider a low-resource translation task from Finnish into Northern Sámi. Collecting all
available parallel data between the languages, we obtain around 30,000 sentence pairs …
available parallel data between the languages, we obtain around 30,000 sentence pairs …
English-Afaan Oromo Machine Translation Using Deep Attention Neural Network
EA Gemechu, GR Kanagachidambaresan - Optical Memory and Neural …, 2023 - Springer
Attention-based neural machine translation (attentional NMT), which jointly aligns and
translates, has got much popularity in recent years. Besides, a language model needs an …
translates, has got much popularity in recent years. Besides, a language model needs an …
Transfer Learning-Based Neural Machine Translation for Low-Resource Languages
J Dong - ACM Transactions on Asian and Low-Resource …, 2023 - dl.acm.org
College of Liberal Arts, Ludong University, Yantai 264025, China Chinese Lexicography
Research Center, Yantai 264025, China Neural Machine Translation (NMT) improves …
Research Center, Yantai 264025, China Neural Machine Translation (NMT) improves …
Improving Morpheme Segmentation Using BERT Embeddings
A Sorokin - International Conference on Analysis of Images, Social …, 2021 - Springer
We offer a method of incorporating BERT embeddings into neural morpheme segmentation.
We show that our method significantly improves over baseline on 6 typologically diverse …
We show that our method significantly improves over baseline on 6 typologically diverse …