DuraS2ST: Chain-of-Thought and Reinforcement Learning for Duration-Aligned Speech-to-Speech Translation

arXiv:2609.33742 · cs.SD, cs.CL, eess.AS · Submitted 2026-09-27 · Read on arXiv

cs.SD, cs.CL, eess.AS

Submitted: 2026-09-27

Updated: 2026-09-27

Code: https://github.com/Mia11939/DuraS2ST

Terminology

• GRPO — 32× in this paper · explained in 6 episodes
• Consistency — 29× in this paper · explained in 3 episodes
• Alignment — 15× in this paper · explained in 1 episode
• Duration Margin Reward — 12× in this paper
• Fidelity — 11× in this paper · explained in 2 episodes
• Reinforcement Learning — 8× in this paper · explained in 4 episodes
• Embeddings — 6× in this paper · explained in 2 episodes
• Modality-Aware Reward Attribution — 4× in this paper
• Accuracy — 2× in this paper · explained in 1 episode
• Adam — 2× in this paper · explained in 2 episodes
• Aggregation — 2× in this paper · explained in 1 episode
• Benchmark — 2× in this paper · explained in 2 episodes
• Core — 2× in this paper · explained in 2 episodes
• Credit assignment — 2× in this paper · explained in 3 episodes
• Transformer — 2× in this paper · explained in 2 episodes
• Unified Framework — 2× in this paper · explained in 5 episodes
• Ablation — 1× in this paper · explained in 1 episode
• Adaptation — 1× in this paper · explained in 1 episode
• Chain-of-Thought (CoT) — 1× in this paper · explained in 4 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Decode — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Group Relative Policy Optimization — 1× in this paper
• Group Relative Policy Optimization (GRPO) — 1× in this paper · explained in 5 episodes
• LoRA Adapters — 1× in this paper · explained in 3 episodes
• Reward Hacking — 1× in this paper · explained in 8 episodes
• Structured Reasoning — 1× in this paper · explained in 3 episodes
• Tokenization — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Seed LiveInterpret 2.0: End-to-end Simultaneous Speech-to-speech Translation with Your Voice
• UniSS: Unified Expressive Speech-to-Speech Translation with Your Voice
• SeamlessM4T: Massively Multilingual & Multimodal Machine Translation
• Seamless: Multilingual Expressive and Streaming Speech Translation
• PolyVoice: Language Models for Speech to Speech Translation
• Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition
• LoRA: Low-Rank Adaptation of Large Language Models
• TranSpeech: Speech-to-Speech Translation With Bilateral Perturbation
• UnitY: Two-pass Direct Speech-to-speech Translation with Discrete Units
• CVSS Corpus and Massively Multilingual Speech-to-Speech Translation
• Kimi-Audio Technical Report
• Efficient Memory Management for Large Language Model Serving with PagedAttention
• TransVIP: Speech to Speech Translation System with Voice and Isochrony Preservation
• Direct speech-to-speech translation with discrete units
• A Call for Clarity in Reporting BLEU Scores
• Robust Speech Recognition via Large-Scale Weak Supervision
• AudioPaLM: A Large Language Model That Can Speak and Listen
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Finetuned Language Models Are Zero-Shot Learners
• Step-Audio 2 Technical Report

Related papers

• Few-Shot Open-Set Audio Classification via Transductive Prototype Refinement and Class Logit Enhancement
• Spectral Masking and Interpolation Attack (SMIA): A Black-box Adversarial Attack against Voice Authentication and Anti-Spoofing Systems
• AVMeme Exam: A Multimodal Multilingual Multicultural Benchmark for LLMs' Contextual and Cultural Knowledge and Thinking
• SoundWeaver: Compositional Warm-Starting for Text-to-Audio Diffusion Serving
• WASIL: In-the-Wild Arabic Spoken Interactions with LLMs
• Efficient Audiovisual Speech Processing via MUTUD: Multimodal Training and Unimodal Deployment