{"title":"增强音频问题解答中的时态理解,建立大型音频语言模型","authors":"Arvind Krishna Sridhar, Yinyi Guo, Erik Visser","doi":"arxiv-2409.06223","DOIUrl":null,"url":null,"abstract":"The Audio Question Answering task includes audio event classification, audio\ncaptioning, and open ended reasoning. Recently, Audio Question Answering has\ngarnered attention due to the advent of Large Audio Language Models. Current\nliterature focuses on constructing LALMs by integrating audio encoders with\ntext only Large Language Models through a projection module. While Large Audio\nLanguage Models excel in general audio understanding, they are limited in\ntemporal reasoning which may hinder their commercial applications and on device\ndeployment. This paper addresses these challenges and limitations in audio\ntemporal reasoning. First, we introduce a data augmentation technique for\ngenerating reliable audio temporal questions and answers using an LLM. Second,\nwe propose a continued finetuning curriculum learning strategy to specialize in\ntemporal reasoning without compromising performance on finetuned tasks.\nFinally, we develop a reliable and transparent automated metric, assisted by an\nLLM, to measure the correlation between Large Audio Language Model responses\nand ground truth data intelligently. We demonstrate the effectiveness of our\nproposed techniques using SOTA LALMs on public audio benchmark datasets.","PeriodicalId":501284,"journal":{"name":"arXiv - EE - Audio and Speech Processing","volume":"56 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Enhancing Temporal Understanding in Audio Question Answering for Large Audio Language Models\",\"authors\":\"Arvind Krishna Sridhar, Yinyi Guo, Erik Visser\",\"doi\":\"arxiv-2409.06223\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The Audio Question Answering task includes audio event classification, audio\\ncaptioning, and open ended reasoning. Recently, Audio Question Answering has\\ngarnered attention due to the advent of Large Audio Language Models. Current\\nliterature focuses on constructing LALMs by integrating audio encoders with\\ntext only Large Language Models through a projection module. While Large Audio\\nLanguage Models excel in general audio understanding, they are limited in\\ntemporal reasoning which may hinder their commercial applications and on device\\ndeployment. This paper addresses these challenges and limitations in audio\\ntemporal reasoning. First, we introduce a data augmentation technique for\\ngenerating reliable audio temporal questions and answers using an LLM. Second,\\nwe propose a continued finetuning curriculum learning strategy to specialize in\\ntemporal reasoning without compromising performance on finetuned tasks.\\nFinally, we develop a reliable and transparent automated metric, assisted by an\\nLLM, to measure the correlation between Large Audio Language Model responses\\nand ground truth data intelligently. We demonstrate the effectiveness of our\\nproposed techniques using SOTA LALMs on public audio benchmark datasets.\",\"PeriodicalId\":501284,\"journal\":{\"name\":\"arXiv - EE - Audio and Speech Processing\",\"volume\":\"56 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - EE - Audio and Speech Processing\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.06223\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - EE - Audio and Speech Processing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.06223","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
音频问题解答任务包括音频事件分类、音频字幕和开放式推理。最近,音频问题解答因大型音频语言模型的出现而备受关注。目前的文献侧重于通过投影模块将音频编码器与纯文本大语言模型集成在一起,从而构建大语言模型。虽然大型音频语言模型在一般音频理解方面表现出色,但在时态推理方面却受到限制,这可能会阻碍其商业应用和设备部署。本文旨在解决音频时态推理中的这些挑战和局限性。首先,我们介绍了一种数据增强技术,利用 LLM 生成可靠的音频时态问题和答案。其次,我们提出了一种持续微调的课程学习策略,在不影响微调任务性能的前提下实现时空推理的专业化。最后,我们在 LLM 的辅助下开发了一种可靠、透明的自动度量方法,用于智能测量大型音频语言模型的回答与地面真实数据之间的相关性。我们在公共音频基准数据集上使用 SOTA LALM 展示了我们提出的技术的有效性。
Enhancing Temporal Understanding in Audio Question Answering for Large Audio Language Models
The Audio Question Answering task includes audio event classification, audio
captioning, and open ended reasoning. Recently, Audio Question Answering has
garnered attention due to the advent of Large Audio Language Models. Current
literature focuses on constructing LALMs by integrating audio encoders with
text only Large Language Models through a projection module. While Large Audio
Language Models excel in general audio understanding, they are limited in
temporal reasoning which may hinder their commercial applications and on device
deployment. This paper addresses these challenges and limitations in audio
temporal reasoning. First, we introduce a data augmentation technique for
generating reliable audio temporal questions and answers using an LLM. Second,
we propose a continued finetuning curriculum learning strategy to specialize in
temporal reasoning without compromising performance on finetuned tasks.
Finally, we develop a reliable and transparent automated metric, assisted by an
LLM, to measure the correlation between Large Audio Language Model responses
and ground truth data intelligently. We demonstrate the effectiveness of our
proposed techniques using SOTA LALMs on public audio benchmark datasets.