Abstract In early 2020, Florida State University Libraries' Office of Digital Research and Scholarship received a small Panhandle Library Access Network (PLAN) Innovation Grant to increase the accessibility of its institutional repository, Diginole, through the creation of video captions and audio transcripts for audiovisual materials. While these audio and video files represent a fairly small percentage of the scholarly outputs found in Diginole, they present a unique set of technical challenges with regard to accessibility when compared to traditional, text-based materials. While text materials can benefit greatly from automatic processes such as OCR (optical character recognition) found in most modern PDF readers and other applications, no such automated, cost-effective “out of the box” solutions are currently available that can guarantee accurate, high-quality captions and transcriptions for AV materials. This column will provide a brief technical history of AV accessibility standards in addition to describing processes of identifying materials for captioning/transcription, selecting a vendor, and provide practical information related to audio and video accessibility that other institutions can hopefully leverage for their own improvements to their IRs.

中文翻译：

---

提高佛罗里达州立大学机构知识库中视听材料的可访问性

摘要 2020 年初，佛罗里达州立大学图书馆数字研究和奖学金办公室获得了一项小型 Panhandle 图书馆访问网络 (PLAN) 创新资助，以通过创建视听视频字幕和音频抄本来提高其机构知识库 Diginole 的可访问性。材料。虽然这些音频和视频文件只占 Diginole 中发现的学术成果的一小部分，但与传统的基于文本的材料相比，它们在可访问性方面提出了一系列独特的技术挑战。虽然文本材料可以从大多数现代 PDF 阅读器和其他应用程序中的 OCR（光学字符识别）等自动化过程中受益匪浅，但没有这种自动化、目前可以使用具有成本效益的“开箱即用”解决方案，可以保证为 AV 材料提供准确、高质量的字幕和转录。除了描述识别字幕/转录材料、选择供应商的过程之外，本专栏还将提供 AV 可访问性标准的简要技术历史，并提供与音频和视频可访问性相关的实用信息，其他机构希望可以利用这些信息来改进自身他们的 IR。