English(EN) CRESiST: Self-Enhancing Explorative Policy Learning for Simultaneous Speech Translation

新框架CRESiST提升同步语音翻译质量和延迟

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-05 05:00

研究人员开发了CRESiST，一个用于同步语音翻译的新颖框架，解决了当前基于LLM的系统中固有的延迟和质量问题。该端到端架构利用了一个全模态LLM和一种交错读写方法，实现了完整的KV缓存重用，并支持无限长度推理，而无需依赖外部语音活动检测。该系统包含一个Actor-Critic管道，用于学习最优流式策略，其中Actor为RNN-T，Critic为LLM，它们使用语义和延迟奖励来提炼最优策略。在各种基准测试上的实验表明，CRESiST建立了新的最先进的质量-延迟帕累托前沿。 AI

影响这项研究可能显著提高实时翻译系统的效率和质量，影响实时字幕和国际通信等应用。

排序理由该集群描述了一篇关于同步语音翻译新颖框架的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 26 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群描述了一篇关于同步语音翻译新颖框架的最新研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Chenyang Le, Pei Zhang, Yuxiang Chen, Bing Han, Derek F. Wong, Baosong Yang, Yanmin Qian · 2026-11-05 05:00

CRESiST：用于同步语音翻译的自增强探索性策略学习

arXiv:2611.06501v1 Announce Type: new Abstract: Simultaneous Speech Translation (SimulST) requires balancing translation quality and latency. While Large Language Models (LLMs) excel in offline translation, current LLM-based SimulST systems predominantly rely on prefix-incrementa…

本文如何被排名

报道来源 [1]

CRESiST：用于同步语音翻译的自增强探索性策略学习

相关实体

相关话题