ABBYY announced the launch of NeoML, an open-source library for building, training, and deploying machine learning models. Available now on GitHub, NeoML supports both deep learning and traditional machine learning algorithms. The cross-platform framework is optimized for applications that run in cloud environments, on desktop and mobile devices. Compared to a popular open-source library (according to internal tests) NeoML offers 15-20% faster performance for pre-trained image processing models. Developers can use NeoML to build, train, and deploy models for object identification, classification, semantic segmentation, verification, and predictive modeling.
NeoML is designed as a universal tool to process and analyze data in a variety of formats including text, image, video, and others. It supports C++, Java, and Objective-C programming languages; Python will be added shortly. NeoML’s neural network models support over 100 layer types. It also offers 20+ traditional ML algorithms such as classification, regression, and clustering frameworks. The library is cross-platform – a single code base that can be run on popular operating systems including Windows, Linux, macOS, iOS, and Android – and optimized for both CPU and GPU processors.
NeoML supports the Open Neural Network Exchange (ONNX), a global open ecosystem for interoperable ML models. The ONNX standard is supported jointly by Microsoft, Facebook, and other partners as an open source project. ABBYY invites developers, data scientists, and business analysts to use and contribute to NeoML on GitHub, where its code is licensed under the Apache License 2.0. The company offers developer support, ongoing review of reports, regular updates, and performance enhancements.
