Posted by Microsoft Lobe is an intuitive open-source platform for developing end-to-end machine learning models. Using Lobe, you can build image classification models and export them into a wide range of formats such as Local API, TensorFlow Lite, TensorFlow, and CoreML to integrate with mobile and web apps.
Lobe fills a sweet spot for customers looking for a simple and quick way to get started with machine learning using their PCs or Macs without requiring any dependency on the cloud, says Microsoft.
The current release, however, can only be used for image classification. But, Microsoft is committed to expanding its capabilities for handling object detection (locate an object inside of an image) and data classification(to label data in a table based on its content).
Microsoft Lobe is available for Windows and iOS platforms. It is free to use and can work without the Internet or login credentials. Privacy of the data you use is ensured for training the machine learning models as the models are trained in the local system.
Images can be imported or captured using the device camera. The images can be labelled in Lobe to train the model. After automatic training, Lobe displays the accuracy, which can be further improved by checking and relabelling the inaccurate classification of images. After evaluating the model, you can export and leverage it with applications; these models can also run on Raspberry Pi projects.
Check out the tour video and visit lobe.ai to download the software.