SALM-Duplex: Efficient and Direct Duplex Modeling for Speech-to-Speech Language Model

• URL: http://arxiv.org/abs/2505.15670v4
• Date: Fri, 25 Jul 2025 15:07:10 GMT
• Title: SALM-Duplex: Efficient and Direct Duplex Modeling for Speech-to-Speech Language Model
• Authors: Ke Hu, Ehsan Hosseini-Asl, Chen Chen, Edresson Casanova, Subhankar Ghosh, Piotr Żelasko, Zhehuai Chen, Jason Li, Jagadeesh Balam, Boris Ginsburg,
• Abstract summary: We propose a novel duplex speech to speech (S2S) architecture featuring continuous user inputs and agent outputs with channel fusion.<n>Using a pretrained streaming for user input enables the first duplex S2S model without requiring speech pretrain.<n> Experimental results show that the proposed model outperforms previous duplex models in reasoning, turn-taking, and barge-in abilities.
• Score: 28.42203609938444
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Spoken dialogue is an intuitive form of human-computer interaction, yet current speech language models often remain constrained to turn-based exchanges, lacking real-time adaptability such as user barge-in. We propose a novel duplex speech to speech (S2S) architecture featuring continuous user inputs and codec agent outputs with channel fusion that directly models simultaneous user and agent streams. Using a pretrained streaming encoder for user input enables the first duplex S2S model without requiring speech pretrain. Separate architectures for agent and user modeling facilitate codec fine-tuning for better agent voices and halve the bitrate (0.6 kbps) compared to previous works. Experimental results show that the proposed model outperforms previous duplex models in reasoning, turn-taking, and barge-in abilities. The model requires significantly less speech data, as speech pretrain is skipped, which markedly simplifies the process of building a duplex S2S model from any LLMs. Finally, it is the first openly available duplex S2S model with training and inference code to foster reproducibility.

Related papers

• TalkingMachines: Real-Time Audio-Driven FaceTime-Style Video via Autoregressive Diffusion Models [2.176487921193175]
  TalkingMachines is an efficient framework that transforms pretrained video generation models into real-time, audio-driven character animators.<n>TalkingMachines enables natural conversational experiences by integrating an audio large language model (LLM) with our video generation foundation model.
  arXiv  Detail & Related papers  (2025-06-03T17:29:28Z)
• VITA-Audio: Fast Interleaved Cross-Modal Token Generation for Efficient Large Speech-Language Model [70.25062476543091]
  VITA-Audio is an end-to-end large speech model with fast audio-text token generation.<n>MCTP module efficiently generates multiple audio tokens within a single model forward pass.<n>Four-stage progressive training strategy is explored to achieve model acceleration with minimal loss of speech quality.
  arXiv  Detail & Related papers  (2025-05-06T17:59:53Z)
• CosyVoice 2: Scalable Streaming Speech Synthesis with Large Language Models [74.80386066714229]
  We present an improved streaming speech synthesis model, CosyVoice 2.<n>Specifically, we introduce finite-scalar quantization to improve codebook utilization of speech tokens.<n>We develop a chunk-aware causal flow matching model to support various synthesis scenarios.
  arXiv  Detail & Related papers  (2024-12-13T12:59:39Z)
• Can We Achieve High-quality Direct Speech-to-Speech Translation without Parallel Speech Data? [49.42189569058647]
  Two-pass direct speech-to-speech translation (S2ST) models decompose the task into speech-to-text translation (S2TT) and text-to-speech (TTS) In this paper, we introduce a composite S2ST model named ComSpeech, which can seamlessly integrate any pretrained S2TT and TTS models into a direct S2ST model. We also propose a novel training method ComSpeech-ZS that solely utilizes S2TT and TTS data.
  arXiv  Detail & Related papers  (2024-06-11T14:17:12Z)
• Cross-Speaker Encoding Network for Multi-Talker Speech Recognition [74.97576062152709]
  Cross-MixSpeaker. Network addresses limitations of SIMO models by aggregating cross-speaker representations. Network is integrated with SOT to leverage both the advantages of SIMO and SISO.
  arXiv  Detail & Related papers  (2024-01-08T16:37:45Z)
• Transfer Learning from Pre-trained Language Models Improves End-to-End Speech Summarization [48.35495352015281]
  End-to-end speech summarization (E2E SSum) directly summarizes input speech into easy-to-read short sentences with a single model. Due to the high cost of collecting speech-summary pairs, an E2E SSum model tends to suffer from training data scarcity and output unnatural sentences. We propose for the first time to integrate a pre-trained language model (LM) into the E2E SSum decoder via transfer learning.
  arXiv  Detail & Related papers  (2023-06-07T08:23:58Z)
• Duplex Diffusion Models Improve Speech-to-Speech Translation [1.4649095013539173]
  Speech-to-speech translation is a sequence-to-sequence learning task that naturally has two directions. We propose a duplex diffusion model that applies diffusion probabilistic models to both sides of a reversible duplex Conformer. Our model enables reversible speech translation by simply flipping the input and output ends.
  arXiv  Detail & Related papers  (2023-05-22T01:39:40Z)
• LAMASSU: Streaming Language-Agnostic Multilingual Speech Recognition and Translation Using Neural Transducers [71.76680102779765]
  Automatic speech recognition (ASR) and speech translation (ST) can both use neural transducers as the model structure. We propose LAMASSU, a streaming language-agnostic multilingual speech recognition and translation model using neural transducers.
  arXiv  Detail & Related papers  (2022-11-05T04:03:55Z)
• Efficiently Fusing Pretrained Acoustic and Linguistic Encoders for Low-resource Speech Recognition [9.732767611907068]
  In this work, we fuse a pre-trained acoustic encoder (wav2vec2.0) and a pre-trained linguistic encoder (BERT) into an end-to-end ASR model. Our model achieves better recognition performance on CALLHOME corpus (15 hours) than other end-to-end models.
  arXiv  Detail & Related papers  (2021-01-17T16:12:44Z)
• Deliberation Model Based Two-Pass End-to-End Speech Recognition [52.45841282906516]
  A two-pass model has been proposed to rescore streamed hypotheses using the non-streaming Listen, Attend and Spell (LAS) model. The model attends to acoustics to rescore hypotheses, as opposed to a class of neural correction models that use only first-pass text hypotheses. A bidirectional encoder is used to extract context information from first-pass hypotheses.
  arXiv  Detail & Related papers  (2020-03-17T22:01:12Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.