Continuous Speech Tokenizer in Text To Speech (2025.findings-naacl)

Copied to clipboard

Challenge: Autoregressive modeling is a common method for processing language sequences and is effective in token prediction.
Approach: They propose a text-to-speech model based on continuous speech tokens and a continuous tokenizer for speech compression.
Outcome: The proposed model has better continuity and higher estimated Mean Opinion Scores (MoS) this is attributed to better information preservation rate across low and high frequencies in the frequency domain.

Similar Papers

Evaluating Text-to-Speech Synthesis from a Large Discrete Token-based Speech Language Model (2024.lrec-main)

Copied to clipboard

Challenge: Recent advances in generative language modeling applied to discrete speech tokens presented a new avenue for text-to-speech (TTS) synthesis.
Approach: They propose to use generative language modeling to generate text-to-speech (TTS) outputs by a discrete token-based model.
Outcome: The proposed model is rated higher in naturalness and context appropriateness in listening tests compared to a conventional TTS.
Generative Spoken Language Model based on continuous word-sized audio tokens (2023.emnlp-main)

Copied to clipboard

Challenge: Text-based language models outperform character-based models, but speech inputs are 20ms or 40ms-long discrete units.
Approach: They propose a generative language model based on word-size continuous audio tokens . they replace lookup table for lexical types with a Lexical Embedding function .
Outcome: The proposed model is five times more memory efficient than discrete unit GSLMs and is phonetically and semantically interpretable.
Speech Discrete Tokens or Continuous Features? A Comparative Analysis for Spoken Language Understanding in SpeechLLMs (2025.emnlp-main)

Copied to clipboard

Challenge: Speech Large Language Models (SpeechLLMs) have emerged as dominant speech processing approaches.
Approach: They compare self-supervised learning-based discrete and continuous features . they compare performance across six spoken language understanding-related tasks .
Outcome: The proposed models outperform discrete tokens and continuous features in six spoken language understanding-related tasks.
RepCodec: A Speech Representation Codec for Speech Tokenization (2024.acl-long)

Copied to clipboard

Challenge: Recent advances in large language models have led to discrete speech tokenization, but this discretization can be costly and impedes performance.
Approach: They propose a new speech representation codec for semantic speech tokenization that reconstructs speech representations from speech encoders like HuBERT or data2vec.
Outcome: The proposed method outperforms the widely used k-means clustering approach in speech understanding and generation.
Language-Codec: Bridging Discrete Codec Representations and Speech Language Models (2025.acl-long)

Copied to clipboard

Challenge: Existing gaps between discrete acoustic codecs and downstream speech language models . initial channel of codebooks contains excessive information, making it difficult to generate tokens from weakly supervised signals such as text.
Approach: They propose a discrete acoustic codec for generating acustic tokens from weakly supervised signals.
Outcome: The proposed language-codec outperforms competing audio compression algorithms and validates on downstream speech language models.
Toward Joint Language Modeling for Speech Units and Text (2023.findings-emnlp)

Copied to clipboard

Challenge: Speech and text are two major forms of human language and little effort has been made to model them together.
Approach: They propose to combine speech and text models to create mixed speech-text data by using different tokenizers and automatic metrics to evaluate how well the model mixes speech and texts.
Outcome: The proposed model improves over a speech-only baseline and shows zero-shot cross-modal transferability.
DiffS2UT: A Semantic Preserving Diffusion Model for Textless Direct Speech-to-Speech Translation (2023.emnlp-main)

Copied to clipboard

Challenge: Existing models for speech generation are not efficient due to low information density of speech data.
Approach: They propose a method to integrate discrete diffusion models into speech generation tasks . they propose to apply diffusion forward process while employing diffusion backward process .
Outcome: The proposed model achieves comparable results to the auto-regressive baselines with significantly fewer decoding steps (50 steps).
SAC: Neural Speech Codec with Semantic-Acoustic Dual-Stream Quantization (2026.acl-long)

Copied to clipboard

Challenge: Existing speech codecs struggle to balance high-quality reconstruction with semantically rich representations, limiting their effectiveness in both generative and understanding tasks.
Approach: They propose a neural speech codec with semantic-acoustic dual-stream quantization that disentangles semantic and acousian modeling into two dedicated streams.
Outcome: The proposed codec outperforms state-of-the-art speech tokenizers in auto-propagating text-to-speech models.
Proteno: Text Normalization with Limited Data for Fast Deployment in Text to Speech Systems (2021.naacl-industry)

Copied to clipboard

Challenge: Developing Text Normalization systems for Text-to-Speech (TTS) on new languages is hard.
Approach: They propose a novel architecture to facilitate Text Normalization systems for TTS on new languages . they use a granular tokenization mechanism that enables the system to learn majority of classes .
Outcome: The proposed architecture performs comparable with the state-of-the-art systems on English . the proposed system learns most classes from training data and precodes them for other classes .
Generative Pre-trained Speech Language Model with Efficient Hierarchical Transformer (2024.acl-long)

Copied to clipboard

Challenge: Experimental results indicate that GPST significantly outperforms the existing speech language models in terms of word error rate, speech quality, and speaker similarity.
Approach: They propose a hierarchical transformer that quantizes audio waveforms into two distinct types of discrete speech representations and integrates them within a transformer architecture.
Outcome: The proposed model outperforms existing speech language models in word error rate, speech quality, and speaker similarity.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations