paper-with-me

홈 › Papers

Robust wav2vec 2.0: Analyzing Domain Shift in Self-Supervised Pre-Training

2021-04-02 · Wei-Ning Hsu, Anuroop Sriram, Alexei Baevski, Tatiana Likhomanenko, Qiantong Xu, Vineel Pratap, Jacob Kahn, Ann Lee, Ronan Collobert, Gabriel Synnaeve, Michael Auli

Self-supervised learning of speech representations has been a very active research area but most work is focused on a single domain such as read audio books for which there exist large quantities of labeled and unlabeled data. In this paper, we explore more general setups where the domain of the unlabeled data for pre-training data differs from the domain of the labeled data for fine-tuning, which in turn may differ from the test data domain. Our experiments show that using target domain data during pre-training leads to large performance improvements across a variety of setups. On a large-scale competitive setup, we show that pre-training on unlabeled in-domain data reduces the gap between models trained on in-domain and out-of-domain labeled data by 66%-73%. This has obvious practical implications since it is much easier to obtain unlabeled target domain data than labeled data. Moreover, we find that pre-training on multiple domains improves generalization performance on domains not seen during training. Code and models will be made available at https://github.com/pytorch/fairseq.

📄 PDF Abstract BibTeX arXiv:2104.01027

Code (3)

pytorch/fairseq 공식 구현 pytorch
hltchkust/cantonese-asr pytorch
marccanby/bidi_decoding pytorch

Tasks

Self-Supervised Learning

Similar Papers 제목 키워드 기반

How Does Pre-trained Wav2Vec 2.0 Perform on Domain Shifted ASR? An Extensive Benchmark on Air Traffic Control Communications

2022-03-31 · Juan Zuluaga-Gomez, Amrutha Prasad, Iuliia Nigmatulina, Saeed Sarfjoo 외

Recent work on self-supervised pre-training focus on leveraging large-scale unlabeled speech data to build robust end-to-end (E2E) acoustic models (AM) that can be later fine-tuned on downstream tasks e.g., automatic spe…

Automatic Speech RecognitionAutomatic Speech Recognition (ASR)speech-recognitionSpeech Recognition

Balanced Learning for Domain Adaptive Semantic Segmentation

2025-12-07 · Wangkai Li, Rui Sun, Bohao Liao, Zhaoyang Li 외 arxiv

Unsupervised domain adaptation (UDA) for semantic segmentation aims to transfer knowledge from a labeled source domain to an unlabeled target domain. Despite the effectiveness of self-training techniques in UDA, they str…

Unsupervised Domain AdaptationSemantic Segmentation

When Does Self-supervision Improve Few-shot Learning?

2019-10-08 · ECCV 2020 8 · Jong-Chyi Su, Subhransu Maji, Bharath Hariharan

We investigate the role of self-supervised learning (SSL) in the context of few-shot learning. Although recent research has shown the benefits of SSL on large unlabeled datasets, its utility on small datasets is relative…

Few-Shot LearningMeta-LearningSelf-Supervised Learning

Mind the Gap: Scanner-induced domain shifts pose challenges for representation learning in histopathology

2022-11-29 · Frauke Wilm, Marco Fragoso, Christof A. Bertram, Nikolas Stathonikos 외

Computer-aided systems in histopathology are often challenged by various sources of domain shift that impact the performance of these algorithms considerably. We investigated the potential of using self-supervised pre-tr…

Representation LearningTumor Segmentation

Complementary Benefits of Contrastive Learning and Self-Training Under Distribution Shift

2023-12-06 · NeurIPS 2023 11 · Saurabh Garg, Amrith Setlur, Zachary Chase Lipton, Sivaraman Balakrishnan 외

Self-training and contrastive learning have emerged as leading techniques for incorporating unlabeled data, both under distribution shift (unsupervised domain adaptation) and when it is absent (semi-supervised learning).…

Contrastive LearningDomain AdaptationUnsupervised Domain Adaptation