谷歌近日宣布开源其深度学习模型Magika,该模型能够以毫秒级的速度准确识别文件格式和内容类型。据谷歌官方博客透露,Magika即使在CPU上运行,也能在几毫秒内精确识别文件类型。在百万文件测试中,Magika的准确率超过了99%。这一开源项目旨在推动人工智能技术的进步和文件识别领域的创新。Magika的源代码已托管至GitHub,供开发者社区使用和进一步优化。
Title: Google Opens Source Magika, AI File Type Identification with 99% Accuracy
Keywords: Google Opens Source, AI Identification, High Accuracy
News Content: Google recently announced the open-sourcing of its deep learning model, Magika, which can accurately identify file formats and content types in milliseconds. According to Google’s official blog, Magika can precisely recognize file types within a few milliseconds even when running on a CPU. During a test with millions of files, Magika achieved an accuracy rate of over 99%. The open-source project aims to advance AI technology and innovation in the field of file identification. The source code of Magika has been hosted on GitHub for the developer community to use and further enhance.
【来源】https://www.ithome.com/0/750/474.htm
Views: 6
