The company is committed to helping people easily create smart applications through a simple visual interface – understanding gestures, listening to music, and reading handwriting without writing any text lines. It is reported that lobe provides windows and MAC desktop clients, which is based on the open source machine learning framework, which helps users train customized machine learning models on their own platform. The current version of lobe can view images (classify the image as a whole into a single label) through classification learning, and Microsoft will try to extend coverage of more types of data in future versions. It is important to note that lobe does not need to be connected or logged in and can only output machine learning models at present. Microsoft team is working on a range of applications and tools to enable users to run the relevant models with little programming. Bill Barnes, lobe project manager, said it used complex technologies to make things more interesting. Hopefully, lobe can bring some confidence in inspiring and practical use of machine learning technology, and then bring more creativity. Global Tech