Voice Memory for Agentic Speech Recognition

  • 类型:arxiv
  • 标识:2607.26410
  • 链接:https://arxiv.org/abs/2607.26410
  • 主分类:agent
  • 形态:method
  • 被引:0
  • 被引来源:Semantic Scholar + OpenAlex
  • S2被引:0
  • OpenAlex被引:0
  • 影响力被引:0
  • TLDR:Voice Memory, a inference-only scheme for agentic speech recognition: at stream time, a frozen corrector reads a single per-domain memory and decides per utterance whether to act on the hypothesis or abstain and keep the 1-best, and transfers across corrector families and adds zero parameters to the inference path.
  • OpenAlex ID:W7171835677
  • OpenAlex DOI:10.48550/arxiv.2607.26410
  • DOI:10.48550/arxiv.2607.26410
  • DOI来源:OpenAlex
  • 开放获取:green
  • 开放获取链接:https://doi.org/10.48550/arxiv.2607.26410
  • OpenAlex更新:2026-08-26
  • 副分类:multimodal
  • 待LLM分类:否
  • 标题中文:Voice Memory:面向 Agentic 语音识别
  • TLDR中文:Voice Memory,一个面向 agentic 语音识别的纯推理方案:流式推理时,冻结 corrector 读取单一 per-domain memory,逐 utterance 决定是否作用于假设或弃权并保留 1-best,跨 corrector 族可迁移,推理路径不增加任何参数。
  • 来源文件
  • /inbox/tom/_candidates/2026-07-31-agent-rag-longcontext-candidates.json
  • [S2 enrich]
  • [OpenAlex backfill]