Investigating the Ability of Large Language Models to Extract Grammatically Correct Sentences from Transcribed Noisy Speech

models 大型语言模型 实验结果 对话 语义理解 语法正确