paper-with-me

홈 › Papers

Automatic Data Augmentation for Domain Adapted Fine-Tuning of Self-Supervised Speech Representations

2023-06-01 · Salah Zaiem, Titouan Parcollet, Slim Essid

Self-Supervised Learning (SSL) has allowed leveraging large amounts of unlabeled speech data to improve the performance of speech recognition models even with small annotated datasets. Despite this, speech SSL representations may fail while facing an acoustic mismatch between the pretraining and target datasets. To address this issue, we propose a novel supervised domain adaptation method, designed for cases exhibiting such a mismatch in acoustic domains. It consists in applying properly calibrated data augmentations on a large clean dataset, bringing it closer to the target domain, and using it as part of an initial fine-tuning stage. Augmentations are automatically selected through the minimization of a conditional-dependence estimator, based on the target dataset. The approach is validated during an oracle experiment with controlled distortions and on two amateur-collected low-resource domains, reaching better performances compared to the baselines in both cases.

📄 PDF Abstract BibTeX arXiv:2306.00481

Code (0)

등록된 구현이 없습니다.

Tasks

Data AugmentationDomain AdaptationSelf-Supervised Learningspeech-recognitionSpeech Recognition

Methods 이 논문이 사용한 방법론

fail 설명 없음

Similar Papers 제목 키워드 기반

PAGANDA: An Adaptive Task-Independent Automatic Data Augmentation

2019-05-16 · ICML Workshop AMTL 2019 6 · Anonymous

Data augmentation is a useful technique to enlarge the size of the training set and prevent overfitting for different machine learning tasks when training data is scarce. However, current data augmentation techniques rel…

BIG-bench Machine LearningData Augmentationimage-classificationImage Classification+1

Adaptive Data Augmentation with Deep Parallel Generative Models

2019-09-25 · Boli Fang, Miao Jiang, Abhirag Nagpure, Jerry Shen

Data augmentation(DA) is a useful technique to enlarge the size of the training set and prevent overfitting for different machine learning tasks when training data is scarce. However, current data augmentation techniques…

BIG-bench Machine LearningData Augmentationimage-classificationImage Classification+1

TADPOLE: Task ADapted Pre-Training via AnOmaLy DEtection

2021-11-01 · EMNLP 2021 11 · Vivek Madan, Ashish Khetan, Zohar Karnin

The paradigm of pre-training followed by finetuning has become a standard procedure for NLP tasks, with a known problem of domain shift between the pre-training and downstream corpus. Previous works have tried to mitigat…

Anomaly DetectionData AugmentationDomain Adaptation

Augmentation-Adapted Retriever Improves Generalization of Language Models as Generic Plug-In

2023-05-27 · Zichun Yu, Chenyan Xiong, Shi Yu, Zhiyuan Liu

Retrieval augmentation can aid language models (LMs) in knowledge-intensive tasks by supplying them with external information. Prior works on retrieval augmentation usually jointly fine-tune the retriever and the LM, mak…

MMLURetrievalZero-shot Generalization

Domain Adaptation of Document-Level NMT in IWSLT19

2019-11-01 · EMNLP (IWSLT) 2019 11 · Martin Popel, Christian Federmann

We describe our four NMT systems submitted to the IWSLT19 shared task in English→Czech text-to-text translation of TED talks. The goal of this study is to understand the interactions between document-level NMT and domain…

Domain AdaptationNMTSentenceTranslation