谷歌近日宣布开源其新型人工智能工具Magika，该工具能够基于人工智能技术快速且精确地识别文件格式和内容类型。据悉，Magika采用了一个定制的、高度优化的深度学习模型，即使在CPU上运行，也能在几毫秒内精确识别文件类型。
这一突破性的技术成果已经过了百万文件的测试，准确率超过99%，表现令人瞩目。谷歌表示，Magika的开源将有助于推动人工智能技术在文件处理和内容识别领域的应用，进一步优化相关产业的工作流程。
英文标题Title：Google Open Sources Magika: Millisecond-Level File Type Recognition
英文关键词Keywords：Google, Magika, File Recognition
英文新闻内容News content：
Google has recently announced the open sourcing of its new artificial intelligence tool Magika, which is capable of rapidly and accurately identifying file formats and content types based on artificial intelligence technology. It is reported that Magika uses a custom-built, highly optimized deep learning model that can precisely recognize file types in just a few milliseconds, even when running on a CPU.
This groundbreaking technological achievement has been tested with millions of files, achieving an accuracy rate of over 99%, which is remarkable. Google stated that the open sourcing of Magika will help to promote the application of artificial intelligence technology in file processing and content recognition, further optimizing the workflows of related industries.
【来源】https://www.ithome.com/0/750/474.htm
Views: 1