Personalization of CTC-based End-to-End Speech Recognition Using Pronunciation-Driven Subword Tokenization

no code implementations • 16 Oct 2023 • Zhihong Lei, Ernest Pusateri, Shiyi Han, Leo Liu, MingBin Xu, Tim Ng, Ruchir Travadi, Youyuan Zhang, Mirko Hannemann, Man-Hung Siu, Zhen Huang

Recent advances in deep learning and automatic speech recognition have improved the accuracy of end-to-end speech recognition systems, but recognition of personal content such as contact names remains a challenge.

Automatic Speech Recognition speech-recognition +1

Acoustic Model Fusion for End-to-end Speech Recognition

no code implementations • 10 Oct 2023 • Zhihong Lei, MingBin Xu, Shiyi Han, Leo Liu, Zhen Huang, Tim Ng, Yuanyuan Zhang, Ernest Pusateri, Mirko Hannemann, Yaqiao Deng, Man-Hung Siu

Recent advances in deep learning and automatic speech recognition (ASR) have enabled the end-to-end (E2E) ASR system and boosted the accuracy to a new level.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +4

Learning from Noisy Labels with Noise Modeling Network

no code implementations • 1 May 2020 • Zhuolin Jiang, Jan Silovsky, Man-Hung Siu, William Hartmann, Herbert Gish, Sancar Adali

Multi-label image classification has generated significant interest in recent years and the performance of such systems often suffers from the not so infrequent occurrence of incorrect or missing labels in the training data.

General Classification Missing Labels +2

Towards a New Understanding of the Training of Neural Networks with Mislabeled Training Data

no code implementations • 18 Sep 2019 • Herbert Gish, Jan Silovsky, Man-Ling Sung, Man-Hung Siu, William Hartmann, Zhuolin Jiang

This includes results about the ability of the noisy model to make the same decisions as the clean model and the effects of noise on model performance.