English
Related papers

Related papers: Balancing Training for Multilingual Neural Machine…

200 papers

Building conversational speech recognition systems for new languages is constrained by the availability of utterances that capture user-device interactions. Data collection is both expensive and limited by the speed of manual transcription.…

Computation and Language · Computer Science 2019-12-03 Surabhi Punjabi , Harish Arsikere , Sri Garimella

The Transformer model has revolutionized Natural Language Processing tasks such as Neural Machine Translation, and many efforts have been made to study the Transformer architecture, which increased its efficiency and accuracy. One potential…

Computation and Language · Computer Science 2023-08-17 Daniela N. Rim , Kimera Richard , Heeyoul Choi

An effective method to improve neural machine translation with monolingual data is to augment the parallel training corpus with back-translations of target language sentences. This work broadens the understanding of back-translation and…

Computation and Language · Computer Science 2018-10-04 Sergey Edunov , Myle Ott , Michael Auli , David Grangier

Machine translation (MT) plays an important role in benefiting linguists, sociologists, computer scientists, etc. by processing natural language to translate it into some other natural language. And this demand has grown exponentially over…

Computation and Language · Computer Science 2019-01-07 Ankush Garg , Mayank Agarwal

We introduce our efforts towards building a universal neural machine translation (NMT) system capable of translating between any language pair. We set a milestone towards this goal by building a single massively multilingual NMT model…

Massively Multilingual Transformer based Language Models have been observed to be surprisingly effective on zero-shot transfer across languages, though the performance varies from language to language depending on the pivot language(s) used…

Computation and Language · Computer Science 2022-05-13 Kabir Ahuja , Shanu Kumar , Sandipan Dandapat , Monojit Choudhury

Neural machine translation has become the state-of-the-art for language pairs with large parallel corpora. However, the quality of machine translation for low-resource languages leaves much to be desired. There are several approaches to…

Computation and Language · Computer Science 2019-10-02 Ilshat Gibadullin , Aidar Valeev , Albina Khusainova , Adil Khan

Training efficiency is one of the main problems for Neural Machine Translation (NMT). Deep networks need for very large data as well as many training iterations to achieve state-of-the-art performance. This results in very high computation…

Computation and Language · Computer Science 2017-10-04 Dakun Zhang , Jungi Kim , Josep Crego , Jean Senellart

Detecting out of policy speech (OOPS) content is important but difficult. While machine learning is a powerful tool to tackle this challenging task, it is hard to break the performance ceiling due to factors like quantity and quality…

Machine Learning · Computer Science 2023-10-25 Apostol Vassilev , Honglan Jin , Munawar Hasan

The multilingual neural machine translation (NMT) model has a promising capability of zero-shot translation, where it could directly translate between language pairs unseen during training. For good transfer performance from supervised…

Computation and Language · Computer Science 2023-05-15 Pengzhi Gao , Liwen Zhang , Zhongjun He , Hua Wu , Haifeng Wang

In this paper, we investigate the problem of training neural machine translation (NMT) systems with a dataset of more than 40 billion bilingual sentence pairs, which is larger than the largest dataset to date by orders of magnitude.…

Computation and Language · Computer Science 2019-10-08 Yuxian Meng , Xiangyuan Ren , Zijun Sun , Xiaoya Li , Arianna Yuan , Fei Wu , Jiwei Li

Neural Machine Translation (NMT) typically leverages monolingual data in training through backtranslation. We investigate an alternative simple method to use monolingual data for NMT training: We combine the scores of a pre-trained and…

Computation and Language · Computer Science 2019-01-25 Felix Stahlberg , James Cross , Veselin Stoyanov

Many valid translations exist for a given sentence, yet machine translation (MT) is trained with a single reference translation, exacerbating data sparsity in low-resource settings. We introduce Simulated Multiple Reference Training (SMRT),…

Computation and Language · Computer Science 2021-04-23 Huda Khayrallah , Brian Thompson , Matt Post , Philipp Koehn

Machine translation systems achieve near human-level performance on some languages, yet their effectiveness strongly relies on the availability of large amounts of parallel sentences, which hinders their applicability to the majority of…

Computation and Language · Computer Science 2018-08-15 Guillaume Lample , Myle Ott , Alexis Conneau , Ludovic Denoyer , Marc'Aurelio Ranzato

The recently proposed massively multilingual neural machine translation (NMT) system has been shown to be capable of translating over 100 languages to and from English within a single model. Its improved translation performance on low…

Computation and Language · Computer Science 2019-11-13 Aditya Siddhant , Melvin Johnson , Henry Tsai , Naveen Arivazhagan , Jason Riesa , Ankur Bapna , Orhan Firat , Karthik Raman

In simultaneous machine translation, the objective is to determine when to produce a partial translation given a continuous stream of source words, with a trade-off between latency and quality. We propose a neural machine translation (NMT)…

Computation and Language · Computer Science 2020-06-01 Patrick Wilken , Tamer Alkhouli , Evgeny Matusov , Pavel Golik

Multimodal machine translation (MMT) aims to improve translation quality by equipping the source sentence with its corresponding image. Despite the promising performance, MMT models still suffer the problem of input degradation: models…

Computer Vision and Pattern Recognition · Computer Science 2022-10-18 Baijun Ji , Tong Zhang , Yicheng Zou , Bojie Hu , Si Shen

In most of neural machine translation distillation or stealing scenarios, the goal is to preserve the performance of the target model (teacher). The highest-scoring hypothesis of the teacher model is commonly used to train a new model…

Computation and Language · Computer Science 2021-04-02 Vilém Zouhar

Modern statistical machine translation (SMT) systems usually use a linear combination of features to model the quality of each translation hypothesis. The linear combination assumes that all the features are in a linear relationship and…

Computation and Language · Computer Science 2015-03-03 Shujian Huang , Huadong Chen , Xinyu Dai , Jiajun Chen

An important concern in training multilingual neural machine translation (NMT) is to translate between language pairs unseen during training, i.e zero-shot translation. Improving this ability kills two birds with one stone by providing an…

Computation and Language · Computer Science 2019-06-21 Ngoc-Quan Pham , Jan Niehues , Thanh-Le Ha , Alex Waibel
‹ Prev 1 3 4 5 6 7 10 Next ›