Blog(6)

• [INTERSPEECH 2025 Series #6] Efficient Streaming TTS Acoustic Model with Depthwise RVQ Decoding Strategies in a Mamba Framework

  Recent advancements in neural codec-based text-to-speech (TTS) systems have revolutionized speech synthesis quality, achieving remarkable naturalness and fidelity.

• MELS-TTS: Multi-Emotion Multi-Lingual Multi-Speaker Text-To-Speech System via Disentangled Style Tokens

  In the swiftly progressing domain of neural text-to-speech (TTS) systems, the quest for creating human-like speech has witnessed remarkable strides. Recent advancements have opened avenues for TTS systems capable of not only mimicking human speech but also encapsulating the nuances of emotions and linguistic diversity.

• Robust neural codec language modeling with phoneme position prediction for zero-shot TTS

  Large language models (LLMs) have exhibited impressive in-context learning abilities [1]. Inspired by these successes, recent studies [2-5] have extended LLM applications to text-to-speech (TTS) systems by representing speech through discrete acoustic codes.

View More

Research Areas(0)

Publications(3)

• Bunched LPCNet2: Efficient Neural Vocoders Covering Devices from Cloud to Edge

  AuthorSangjun Park, Kihyun Choo, Joohyung Lee, Anton V. Porov, Konstantin Osipov, June Sig Sung

  PublishedAnnual Conference of the International Speech Communication Association (INTERSPEECH)

  Date2022-09-18

• Part-of-Speech Models Compression Methods for On-device Grapheme-to-phoneme Conversion

  AuthorMarek Kubis, Pawel Skorzewski, Marcin Lewandowski 외

  PublishedInternational Conference on Acoustics, Speech, and Signal Processing (ICASSP)

  Date2022-05-22

• SEMI-SUPERVISED TRANSFER LEARNING FOR LANGUAGE EXPANSION OF END-TO-END SPEECH RECOGNITION MODELS TO LOW-RESOURCE LANGUAGES

  AuthorJiyeon Kim, Mehul Kumar, Dhananjaya Gowda, Abhinav Garg, Chanwoo Kim

  PublishedIEEE Workshop on Automatic Speech Recognition & Understanding (ASRU)

  Date2021-12-17

News(9)

• [Blog] Robust neural codec language modeling with phoneme position prediction for zero-shot TTS

  Large language models (LLMs) have exhibited impressive in-context learning abilities [1]. Inspired by these successes, recent studies [2-5] have extended LLM applications to text-to-speech (TTS) systems by representing speech through discrete acoustic codes.

• [Blog] Fairness in Dysarthric Speech Synthesis: Understanding Intrinsic Bias in Dysarthric Speech Cloning using F5-TTS

  Voice cloning, especially zero-shot speech synthesis, has become one of the most exciting frontiers in speech technology.

• [Blog] Efficient Streaming TTS Acoustic Model with Depthwise RVQ Decoding Strategies in a Mamba Framework

  Recent advancements in neural codec-based text-to-speech (TTS) systems have revolutionized speech synthesis quality, achieving remarkable naturalness and fidelity.

View More

Others(0)