dlib

dlib is a long-running open-source C++ library with machine learning, deep learning and image processing algorithms, plus Python bindings. Developers use it widely for face detection, facial landmark tracking and object detection in their own software.

The library covers support vector machines, kernel methods, Bayesian networks, clustering, optimization and a deep learning API for building and training networks. Pretrained examples include face recognition, landmark detection, vehicle detection and semantic segmentation.

Beyond machine learning, dlib ships general utilities such as linear algebra, networking, an HTTP server, compression and FFmpeg-based video capture. The documentation is organized around many example programs, and the code is free to use under the Boost Software License.