Microsoft announced that Lobe, a machine learning tool, can now provide image classification support. The machine can help people quickly apply deep learning and AI models to the tools they are developing without writing code. Essentially, this means that people can import images of the things they want Lobe to recognize, and the free app will automatically select the correct machine learning architecture to start training machine learning models. The company said that it will publicly preview the application today, and it can be downloaded for free on Windows or Mac computers.
“Today, Lobe supports image classification, but plans to expand to other models and data types in the future,” said Jennifer Langston, who wrote articles about Microsoft research and innovation, in a blog. Lobe’s visual interface allows developers to create applications with functions such as reading handwriting, recognizing gestures, and listening to music. You can download Lobe for Windows and Mac from the Lobe page.
Microsoft acquired Lobe, a company focused on AI in 2018, as part of its strategy and CEO Satya Nadella’s vision to integrate AI into all products the company produces. Lobe helps developers and those without any data science experience quickly apply deep learning and AI models to tools that can be used in industries such as healthcare or agriculture. Lobe uses the AI research, global infrastructure and developer tools of the tech giant.
Microsoft said that Lobe uses “open source machine learning architecture and transfer learning to train custom machine learning models on users’ own machines.” The company claims that the data is confidential and does not require an internet connection or login. With image classification support, people can simply import images of things they want to train Lobe, and the training is done automatically. In addition, it complements Azure AI’s services for customers who want to take advantage of cloud computing capabilities.