luigifreda/slamplay
A C++ SLAM development framework integrating deep learning, optimization, and computer vision libraries with examples for visual and LiDAR-inertial SLAM.

slamplay provides a unified CMake framework that wires together major SLAM building blocks including back-end optimizers (g2o, GTSAM, Ceres), front-end perception tools (OpenCV, PCL), deep learning runtimes (TensorRT, TensorFlow, libtorch), and visualization libraries (Pangolin, ImGui, ReRun). It includes commented examples and supports visual SLAM and LiDAR-inertial SLAM pipelines, with deep learning modules for depth estimation, semantic segmentation, and object detection.