A Domain Adaptation Framework for Speech Recognition Systems with Only Synthetic data

no code implementations • 21 Jan 2025 • Minh Tran, Yutong Pang, Debjyoti Paul, Laxmi Pandey, Kevin Jiang, Jinxi Guo, Ke Li, Shun Zhang, Xuedong Zhang, Xin Lei

We introduce DAS (Domain Adaptation with Synthetic data), a novel domain adaptation framework for pre-trained ASR model, designed to efficiently adapt to various language-defined domains without requiring any real data.

Domain Adaptation speech-recognition +3

LLaMA based Punctuation Restoration With Forward Pass Only Decoding

no code implementations • 9 Aug 2024 • Yutong Pang, Debjyoti Paul, Kevin Jiang, Xuedong Zhang, Xin Lei

This paper introduces two advancements in the field of Large Language Model Annotation with a focus on punctuation restoration tasks.

Language Modeling Language Modelling +2

Towards scalable efficient on-device ASR with transfer learning

no code implementations • 23 Jul 2024 • Laxmi Pandey, Ke Li, Jinxi Guo, Debjyoti Paul, Arthur Guo, Jay Mahadeokar, Xuedong Zhang

Multilingual pretraining for transfer learning significantly boosts the robustness of low-resource monolingual ASR models.

Transfer Learning

SignDiff: Diffusion Models for American Sign Language Production

no code implementations • 30 Aug 2023 • Sen Fang, Chunyu Sui, Yanghao Zhou, Xuedong Zhang, Hongbin Zhong, Minyu Zhao, Yapeng Tian, Chen Chen

In this paper, we propose a dual-condition diffusion pre-training model named SignDiff that can generate human sign language speakers from a skeleton pose.

Pose Estimation Sign Language Production +1

Language Agnostic Data-Driven Inverse Text Normalization

no code implementations • 20 Jan 2023 • Szu-Jui Chen, Debjyoti Paul, Yutong Pang, Peng Su, Xuedong Zhang

With the emergence of automatic speech recognition (ASR) models, converting the spoken form text (from ASR) to the written form is in urgent need.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +3

Adaptive Multi-Corpora Language Model Training for Speech Recognition

no code implementations • 9 Nov 2022 • Yingyi Ma, Zhe Liu, Xuedong Zhang

Thus, the data sampling strategy is important to the adaptation performance.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +4

Mitigating Unintended Memorization in Language Models via Alternating Teaching

no code implementations • 13 Oct 2022 • Zhe Liu, Xuedong Zhang, Fuchun Peng

Recent research has shown that language models have a tendency to memorize rare or unique sequences in the training corpora which can thus leak sensitive attributes of user data.

Memorization Privacy Preserving

Improving Data Driven Inverse Text Normalization using Data Augmentation

no code implementations • 20 Jul 2022 • Laxmi Pandey, Debjyoti Paul, Pooja Chitkara, Yutong Pang, Xuedong Zhang, Kjell Schubert, Mark Chou, Shu Liu, Yatharth Saraf

Inverse text normalization (ITN) is used to convert the spoken form output of an automatic speech recognition (ASR) system to a written form.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +3

Speech recognition for medical conversations

no code implementations • 20 Nov 2017 • Chung-Cheng Chiu, Anshuman Tripathi, Katherine Chou, Chris Co, Navdeep Jaitly, Diana Jaunzeikare, Anjuli Kannan, Patrick Nguyen, Hasim Sak, Ananth Sankar, Justin Tansuwan, Nathan Wan, Yonghui Wu, Xuedong Zhang

We explored both CTC and LAS systems for building speech recognition models.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +1