7 papers
Latent Softmax for Data-Efficient Phoneme-Based Multilingual ASR Across Tonal and Non-Tonal Languages
Saierdaer Yusuyin, Nanling Jiang, Hao Huang +1
Phoneme-based multilingual automatic speech recognition (ASR) can share acoustic evidence across languages more directly than language-specific subword modeling. When tonal and non…
Phonemes vs. Projectors: An Investigation of Speech-Language Interfaces for LLM-based ASR
Ziwei Li, Lukuang Dong, Saierdaer Yusuyin +2
Integrating pretrained speech encoders with large language models (LLMs) is promising for ASR, but performance and data efficiency depend on the speech-language interface. A common…
Advancing LLM-based phoneme-to-grapheme for multilingual speech recognition
Lukuang Dong, Ziwei Li, Saierdaer Yusuyin +2
Phoneme-based ASR factorizes recognition into speech-to-phoneme (S2P) and phoneme-to-grapheme (P2G), enabling cross-lingual acoustic sharing while keeping language-specific orthogr…
CTC-TTS: LLM-based dual-streaming text-to-speech with CTC alignment
Hanwen Liu, Saierdaer Yusuyin, Hao Huang +1
Large-language-model (LLM)-based text-to-speech (TTS) systems can generate natural speech, but most are not designed for low-latency dual-streaming synthesis. High-quality dual-str…
Pronunciation-Lexicon Free Training for Phoneme-based Crosslingual ASR via Joint Stochastic Approximation
Saierdaer Yusuyin, Te Ma, Hao Huang +1
Recently, pre-trained models with phonetic supervision have demonstrated their advantages for crosslingual speech recognition in data efficiency and information sharing across lang…
LLM-based phoneme-to-grapheme for phoneme-based speech recognition
Te Ma, Min Bi, Saierdaer Yusuyin +2
In automatic speech recognition (ASR), phoneme-based multilingual pre-training and crosslingual fine-tuning is attractive for its high data efficiency and competitive results compa…