Dlib

Dlib is the high-performance C++ toolkit for real-world machine learning (ML), computer vision, and data analysis applications.

Dlib, a robust, cross-platform C++ library, delivers a comprehensive suite of tools for solving complex problems in ML and data analysis. Developed by Davis E. King since 2002, it is used across industry and academia (robotics, embedded devices, HPC) under the permissive Boost Software License. Key features include state-of-the-art algorithms: structural Support Vector Machines (SVMs), deep learning (CNNs for object detection), and efficient image processing (HOG-based face detection). The library emphasizes high-quality, portable code, complete documentation, and robust Python bindings, making it a reliable choice for production systems.

https://dlib.net