Freecodec: A Disentangled Neural Speech Codec With Fewer Tokens

Abstract

Neural speech codecs have gained great attention for their outstanding reconstruction with discrete token representations. It is a crucial component in generative tasks such as speech coding and large language models (LLM). However, most works based on residual vector quantization perform worse with fewer tokens due to low coding efficiency for modeling complex coupled information. In this paper, we propose a neural speech codec named FreeCodec which employs a more effective encoding framework by decomposing intrinsic properties of speech into different components: 1) a global vector is extracted as the timbre information, 2) a prosody encoder with a long stride level is used to model the prosody information, 3) the content information is from a content encoder. Using different training strategies, FreeCodec achieves state-of-the-art performance in reconstruction and disentanglement scenarios. Results from subjective and objective experiments demonstrate that our framew

Stats

• citations3
• S2 citations—
• github stars0
• HF likes0
• heat score4.52
• arxiv keyzheng2024freecodec

Related papers

• Funcodec: A Fundamental, Reproducible And Integrable Open-source Toolkit For Neural Speech Codec (2023)17.47
• ESC: Efficient Speech Coding With Cross-scale Residual Vector Quantized Transformers (2024)5.84
• Disentangled Feature Learning For Real-time Neural Speech Coding (2022)0.00
• Optimizing Neural Speech Codec For Low-bitrate Compression Via Multi-scale Encoding (2024)0.00
• Lscodec: Low-bitrate And Speaker-decoupled Discrete Speech Codec (2024)0.00
• Fewer-token Neural Speech Codec With Time-invariant Codes (2023)10.07
• Disco-speech: Controllable Zero-shot Speech Generation With A Disentangled Speech Codec (2025)0.00
• Msr-codec: A Low-bitrate Multi-stream Residual Codec For High-fidelity Speech Generation With Information Disentanglement (2025)2.35