monocular-slam

Educational/research monocular visual odometry implementation with ORB features, initialization, tracking, local mapping, and bundle adjustment in C++.

[3DV 2026] ViSTA-SLAM: Visual SLAM with Symmetric Two-view Association

[3DV'24 Best Paper Honorable Mention] NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM

A ROS wrapper for ORB-SLAM2

Real-Time SLAM for Monocular, Stereo and RGB-D Cameras, with Loop Detection and Relocalization Capabilities

One RGB camera, 8.2 mm from motion capture - and no idea how big the room is. Reproducible monocular SfM on TUM fr3, plus the tool that shows what scale alignment is hiding.

Dive into cutting-edge FusionSLAM, where SuperPoint, SuperGlue, Neural Depth Estimation, and Instant-NGP converge, elevating Monocular SLAM to unparalleled precision and performance. Redefining mapping, localization, and reconstruction in a single camera setup.

Real-time SLAM using ORB-SLAM3 with Python bindings

SLAM system using Total Least Square algorithm to determine the trajectory of a robot moving in a given environment

An implementation of Graph-based SLAM using only an onboard monocular camera. Developed as part of MSc Robotics Masters Thesis (2017) at University of Birmingham.

MonoSLAM to auto calibrate endoscope cameras

Landmark detection and localization project using python.

Source code for virtual-based experiments presented in the paper: UAV Navigation Using EKF-MonoSLAM Aided by Range- to-Base Measurements

Off-device monocular dense VSLAM benchmarking framework for uncalibrated smartphone video streams, evaluating trajectory accuracy, dense 3D reconstruction quality, and efficiency on ADVIO, TUM-RGBD & custom Record3D datasets. Real-time concurrent Ray streaming pipeline.

Replicated results from DenseDepth using DenseNet169 in Python.

ROS2 visual SLAM with VGGT: monocular dense 3D mapping, sliding-window inference, TF2/Nav2, Docker + CUDA.

Camera trajectory and a sparse 3D point cloud from a single-lens RGB video. Classical multi-view geometry, no neural network, no GPU: a 10 second clip reconstructs in under 7 seconds on CPU.

Monocular visual SLAM with KLT/SIFT tracking, local bundle adjustment, SIFT loop closure + Sim(3) pose-graph optimization. ETH VAMR course project.

Monocular SLAM triangle splatting ML based feature extraction

🔍 Explore visual SLAM with ViSTA-SLAM, utilizing symmetric two-view association for accurate mapping and navigation in diverse environments.