# Noise Adaptive Streaming Audio-Visual Speech Token Enhancement for Robust Full-Duplex Spoken Dialogue Models

arXiv: https://arxiv.org/abs/2609.08390 · HTML: https://mlsanity.com/paper/2609.08390/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 22 of its 29 references that are in the map.
- First posted on arXiv 8 Sep 2026.

## Builds on (most cited first, 22 of 22)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [HuBERT: Self-Supervised Speech Representation Learning by Masked Prediction of Hidden Units](https://mlsanity.com/paper/2106.07447/) (2021)
- [SoundStream: An End-to-End Neural Audio Codec](https://mlsanity.com/paper/2107.03312/) (2021)
- [Moshi: a speech-text foundation model for real-time dialogue](https://mlsanity.com/paper/2410.00037/) (2024)
- [LRS3-TED: a large-scale dataset for visual speech recognition](https://mlsanity.com/paper/1809.00496/) (2018)
- [Learning Audio-Visual Speech Representation by Masked Multimodal Cluster Prediction](https://mlsanity.com/paper/2201.02184/) (2022)
- [GLM-4-Voice: Towards Intelligent and Human-Like End-to-End Spoken Chatbot](https://mlsanity.com/paper/2412.02612/) (2024)
- [Freeze-Omni: A Smart and Low Latency Speech-to-speech Dialogue Model with Frozen LLM](https://mlsanity.com/paper/2411.00774/) (2024)
- [Auto-AVSR: Audio-Visual Speech Recognition with Automatic Labels](https://mlsanity.com/paper/2303.14307/) (2023)
- [Emformer: Efficient Memory Transformer Based Acoustic Model For Low Latency Streaming Speech Recognition](https://mlsanity.com/paper/2010.10759/) (2020)
- [Full-Duplex-Bench: A Benchmark to Evaluate Full-duplex Spoken Dialogue Models on Turn-taking Capabilities](https://mlsanity.com/paper/2503.04721/) (2025)
- [Streaming automatic speech recognition with the transformer model](https://mlsanity.com/paper/2001.02674/) (2020)
- [Beyond Turn-Based Interfaces: Synchronous LLMs as Full-Duplex Dialogue Agents](https://mlsanity.com/paper/2409.15594/) (2024)
- [PersonaPlex: Voice and Role Control for Full Duplex Conversational Speech Models](https://mlsanity.com/paper/2602.06053/) (2026)
- [Canary-1B-v2 & Parakeet-TDT-0.6B-v3: Efficient and High-Performance Models for Multilingual ASR and AST](https://mlsanity.com/paper/2509.14128/) (2025)
- [Where Visual Speech Meets Language: VSP-LLM Framework for Efficient and Context-Aware Visual Speech Processing](https://mlsanity.com/paper/2402.15151/) (2024)
- [SALMONN-omni: A Standalone Speech LLM without Codec Injection for Full-duplex Conversation](https://mlsanity.com/paper/2505.17060/) (2025)
- [MMS-LLaMA: Efficient LLM-based Audio-Visual Speech Recognition with Minimal Multimodal Speech Tokens](https://mlsanity.com/paper/2503.11315/) (2025)
- [AV-Dialog: Spoken Dialogue Models with Audio-Visual Input](https://mlsanity.com/paper/2511.11124/) (2025)
- [PASE: Leveraging the Phonological Prior of WavLM for Low-Hallucination Generative Speech Enhancement](https://mlsanity.com/paper/2511.13300/) (2025)
- [Vision-Speech Models: Teaching Speech Models to Converse about Images](https://mlsanity.com/paper/2503.15633/) (2025)

## Built on by (most cited first, 1 of 1)

- [Long-Term Memory-Guided Enhancement for Target Perception in Audio-Language Models](https://mlsanity.com/paper/2609.36577/) (2026)

## Abstract (arXiv)

Full-duplex spoken dialogue systems enable simultaneous listening and speaking, but their audio-only perception often fails under background noise and overlapping speech, leading to incoherent responses. Recent audio-visual dialogue approaches show that incorporating visual cues such as lip movements improve robustness under audio corruption. However, existing approaches often adapt the large speech dialogue model itself to process visual input, requiring costly multimodal training. We propose AV-STE, a modular streaming audio-visual front-end that restores corrupted semantic speech tokens from noisy audio and lip video before they reach the speech LLM. The downstream dialogue model remains entirely frozen, preserving its pretrained conversational capabilities. When integrated with frozen Moshi, AV-STE improves average GPT-4o-judged response coherence from 1.42 to 1.91 under same-dataset speaker interference while largely preserving turn-taking behavior. Gains also transfer to out-of-domain Seamless Interaction.