scholarly journals Text-to-speech system for low-resource language using cross-lingual transfer learning and data augmentation

2021 ◽  
Vol 2021 (1) ◽  
Author(s):  
Zolzaya Byambadorj ◽  
Ryota Nishimura ◽  
Altangerel Ayush ◽  
Kengo Ohta ◽  
Norihide Kitaoka
Keyword(s):  
Transfer Learning ◽  
Data Augmentation ◽  
Prediction Models ◽  
Target Language ◽  
Text To Speech ◽  
Paired Data ◽  
Low Resource ◽  
Language Data ◽  
Single Speaker ◽  
Cross Lingual

AbstractDeep learning techniques are currently being applied in automated text-to-speech (TTS) systems, resulting in significant improvements in performance. However, these methods require large amounts of text-speech paired data for model training, and collecting this data is costly. Therefore, in this paper, we propose a single-speaker TTS system containing both a spectrogram prediction network and a neural vocoder for the target language, using only 30 min of target language text-speech paired data for training. We evaluate three approaches for training the spectrogram prediction models of our TTS system, which produce mel-spectrograms from the input phoneme sequence: (1) cross-lingual transfer learning, (2) data augmentation, and (3) a combination of the previous two methods. In the cross-lingual transfer learning method, we used two high-resource language datasets, English (24 h) and Japanese (10 h). We also used 30 min of target language data for training in all three approaches, and for generating the augmented data used for training in methods 2 and 3. We found that using both cross-lingual transfer learning and augmented data during training resulted in the most natural synthesized target speech output. We also compare single-speaker and multi-speaker training methods, using sequential and simultaneous training, respectively. The multi-speaker models were found to be more effective for constructing a single-speaker, low-resource TTS model. In addition, we trained two Parallel WaveGAN (PWG) neural vocoders, one using 13 h of our augmented data with 30 min of target language data and one using the entire 12 h of the original target language dataset. Our subjective AB preference test indicated that the neural vocoder trained with augmented data achieved almost the same perceived speech quality as the vocoder trained with the entire target language dataset. Overall, we found that our proposed TTS system consisting of a spectrogram prediction network and a PWG neural vocoder was able to achieve reasonable performance using only 30 min of target language training data. We also found that by using 3 h of target language data, for training the model and for generating augmented data, our proposed TTS model was able to achieve performance very similar to that of the baseline model, which was trained with 12 h of target language data.

scholarly journals Cross-lingual transfer of sentiment classifiers

2021 ◽  
Vol 9 (1) ◽  
pp. 1-25
Author(s):  
Marko Robnik-Šikonja ◽  
Kristjan Reba ◽  
Igor Mozetič
Keyword(s):  
Machine Learning ◽  
Vector Space ◽  
Prediction Models ◽  
Language Models ◽  
Target Language ◽  
Word Embeddings ◽  
Language Data ◽  
Cross Lingual ◽  
Multiple Languages ◽  
Transfer Mechanisms

Word embeddings represent words in a numeric space so that semantic relations between words are represented as distances and directions in the vector space. Cross-lingual word embeddings transform vector spaces of different languages so that similar words are aligned. This is done by mapping one language’s vector space to the vector space of another language or by construction of a joint vector space for multiple languages. Cross-lingual embeddings can be used to transfer machine learning models between languages, thereby compensating for insufficient data in less-resourced languages. We use cross-lingual word embeddings to transfer machine learning prediction models for Twitter sentiment between 13 languages. We focus on two transfer mechanisms that recently show superior transfer performance. The first mechanism uses the trained models whose input is the joint numerical space for many languages as implemented in the LASER library. The second mechanism uses large pretrained multilingual BERT language models. Our experiments show that the transfer of models between similar languages is sensible, even with no target language data. The performance of cross-lingual models obtained with the multilingual BERT and LASER library is comparable, and the differences are language-dependent. The transfer with CroSloEngual BERT, pretrained on only three languages, is superior on these and some closely related languages.

scholarly journals A Joint Back-Translation and Transfer Learning Method for Low-Resource Neural Machine Translation

2020 ◽  
Vol 2020 ◽  
pp. 1-11
Author(s):  
Gong-Xu Luo ◽  
Ya-Ting Yang ◽  
Rui Dong ◽  
Yan-Hong Chen ◽  
Wen-Bo Zhang
Keyword(s):  
Machine Translation ◽  
Transfer Learning ◽  
Large Scale ◽  
Data Augmentation ◽  
Training Methods ◽  
Learning Method ◽  
Neural Machine Translation ◽  
Low Resource ◽  
Parallel Data ◽  
Back Translation

Neural machine translation (NMT) for low-resource languages has drawn great attention in recent years. In this paper, we propose a joint back-translation and transfer learning method for low-resource languages. It is widely recognized that data augmentation methods and transfer learning methods are both straight forward and effective ways for low-resource problems. However, existing methods, which utilize one of these methods alone, limit the capacity of NMT models for low-resource problems. In order to make full use of the advantages of existing methods and further improve the translation performance of low-resource languages, we propose a new method to perfectly integrate the back-translation method with mainstream transfer learning architectures, which can not only initialize the NMT model by transferring parameters of the pretrained models, but also generate synthetic parallel data by translating large-scale monolingual data of the target side to boost the fluency of translations. We conduct experiments to explore the effectiveness of the joint method by incorporating back-translation into the parent-child and the hierarchical transfer learning architecture. In addition, different preprocessing and training methods are explored to get better performance. Experimental results on Uygur-Chinese and Turkish-English translation demonstrate the superiority of the proposed method over the baselines that use single methods.

Deep Persian sentiment analysis: Cross-lingual training for low-resource languages

2020 ◽  
pp. 016555152096278
Author(s):  
Rouzbeh Ghasemi ◽  
Seyed Arad Ashrafi Asli ◽  
Saeedeh Momtazi
Keyword(s):  
Natural Language Processing ◽  
Natural Language ◽  
Sentiment Analysis ◽  
Language Processing ◽  
Training Data ◽  
Target Language ◽  
Low Resource ◽  
Proposed Model ◽  
Significant Difference ◽  
Cross Lingual

With the advent of deep neural models in natural language processing tasks, having a large amount of training data plays an essential role in achieving accurate models. Creating valid training data, however, is a challenging issue in many low-resource languages. This problem results in a significant difference between the accuracy of available natural language processing tools for low-resource languages compared with rich languages. To address this problem in the sentiment analysis task in the Persian language, we propose a cross-lingual deep learning framework to benefit from available training data of English. We deployed cross-lingual embedding to model sentiment analysis as a transfer learning model which transfers a model from a rich-resource language to low-resource ones. Our model is flexible to use any cross-lingual word embedding model and any deep architecture for text classification. Our experiments on English Amazon dataset and Persian Digikala dataset using two different embedding models and four different classification networks show the superiority of the proposed model compared with the state-of-the-art monolingual techniques. Based on our experiment, the performance of Persian sentiment analysis improves 22% in static embedding and 9% in dynamic embedding. Our proposed model is general and language-independent; that is, it can be used for any low-resource language, once a cross-lingual embedding is available for the source–target language pair. Moreover, by benefitting from word-aligned cross-lingual embedding, the only required data for a reliable cross-lingual embedding is a bilingual dictionary that is available between almost all languages and the English language, as a potential source language.

scholarly journals Cross-Lingual Pre-Training Based Transfer for Zero-Shot Neural Machine Translation

2020 ◽  
Vol 34 (01) ◽  
pp. 115-122 ◽  
Author(s):  
Baijun Ji ◽  
Zhirui Zhang ◽  
Xiangyu Duan ◽  
Min Zhang ◽  
Boxing Chen ◽  
...  
Keyword(s):  
Machine Translation ◽  
Transfer Learning ◽  
Large Scale ◽  
Feature Space ◽  
Target Language ◽  
Smooth Transition ◽  
Training Methods ◽  
Neural Machine Translation ◽  
Cross Lingual ◽  
Effective Transfer

Transfer learning between different language pairs has shown its effectiveness for Neural Machine Translation (NMT) in low-resource scenario. However, existing transfer methods involving a common target language are far from success in the extreme scenario of zero-shot translation, due to the language space mismatch problem between transferor (the parent model) and transferee (the child model) on the source side. To address this challenge, we propose an effective transfer learning approach based on cross-lingual pre-training. Our key idea is to make all source languages share the same feature space and thus enable a smooth transition for zero-shot translation. To this end, we introduce one monolingual pre-training method and two bilingual pre-training methods to obtain a universal encoder for different languages. Once the universal encoder is constructed, the parent model built on such encoder is trained with large-scale annotated data and then directly applied in zero-shot translation scenario. Experiments on two public datasets show that our approach significantly outperforms strong pivot-based baseline and various multilingual NMT approaches.

scholarly journals Improving Candidate Generation for Low-resource Cross-lingual Entity Linking

2020 ◽  
Vol 8 ◽  
pp. 109-124
Author(s):  
Shuyan Zhou ◽  
Shruti Rijhwani ◽  
John Wieting ◽  
Jaime Carbonell ◽  
Graham Neubig
Keyword(s):  
State Of The Art ◽  
Target Language ◽  
Entity Linking ◽  
Average Gain ◽  
Source Language ◽  
Low Resource ◽  
High Resource ◽  
Language Knowledge ◽  
Cross Lingual ◽  
Improved Model

Cross-lingual entity linking (XEL) is the task of finding referents in a target-language knowledge base (KB) for mentions extracted from source-language texts. The first step of (X)EL is candidate generation, which retrieves a list of plausible candidate entities from the target-language KB for each mention. Approaches based on resources from Wikipedia have proven successful in the realm of relatively high-resource languages, but these do not extend well to low-resource languages with few, if any, Wikipedia pages. Recently, transfer learning methods have been shown to reduce the demand for resources in the low-resource languages by utilizing resources in closely related languages, but the performance still lags far behind their high-resource counterparts. In this paper, we first assess the problems faced by current entity candidate generation methods for low-resource XEL, then propose three improvements that (1) reduce the disconnect between entity mentions and KB entries, and (2) improve the robustness of the model to low-resource scenarios. The methods are simple, but effective: We experiment with our approach on seven XEL datasets and find that they yield an average gain of 16.9% in Top-30 gold candidate recall, compared with state-of-the-art baselines. Our improved model also yields an average gain of 7.9% in in-KB accuracy of end-to-end XEL. 1

scholarly journals An Evaluation of Multilingual Offensive Language Identification Methods for the Languages of India

Information ◽  
2021 ◽  
Vol 12 (8) ◽  
pp. 306
Author(s):  
Tharindu Ranasinghe ◽  
Marcos Zampieri
Keyword(s):  
Computational Models ◽  
Data Augmentation ◽  
New Technology ◽  
Language Identification ◽  
Low Resource ◽  
High Resource ◽  
Augmentation Strategies ◽  
Cross Lingual ◽  
Offensive Language ◽  
Better Than

The pervasiveness of offensive content in social media has become an important reason for concern for online platforms. With the aim of improving online safety, a large number of studies applying computational models to identify such content have been published in the last few years, with promising results. The majority of these studies, however, deal with high-resource languages such as English due to the availability of datasets in these languages. Recent work has addressed offensive language identification from a low-resource perspective, exploring data augmentation strategies and trying to take advantage of existing multilingual pretrained models to cope with data scarcity in low-resource scenarios. In this work, we revisit the problem of low-resource offensive language identification by evaluating the performance of multilingual transformers in offensive language identification for languages spoken in India. We investigate languages from different families such as Indo-Aryan (e.g., Bengali, Hindi, and Urdu) and Dravidian (e.g., Tamil, Malayalam, and Kannada), creating important new technology for these languages. The results show that multilingual offensive language identification models perform better than monolingual models and that cross-lingual transformers show strong zero-shot and few-shot performance across languages.

Sign in / Sign up

Export Citation Format

Share Document