Benchmark for UAV Visual Localization under Low-altitude Multi-view Observation Condition
A robust vision-based localization system for UAVs operating in GPS-denied environments. The system utilizes advanced computer vision techniques to achieve accurate position estimation by matching real-time drone camera feeds with satellite imagery.
Official code for VoLN: Vision-Only Long-Horizon Navigation—Paradigm, Benchmark, and Method. An embodied AI UAV benchmark and VoLN-MLLM agent bridging VLN and multimodal LLMs across 7,210 episodes, AirSim, and real-world flights.
[ACM MM 2026] Official implementation of Self-in-Space: Benchmarking Self-Awareness and Spatial Cognition in UAV Embodied Intelligence
Scale-Aware UAV-to-Satellite Cross-View Geo-Localization
Visual-intertial-odometry system for UAVs
6-DOF free-flight drone simulator built on Luanti. Features 5 camera views, 1st/3rd person toggle, HUD telemetry, minimap navigation, and a foundation for visual RL & VLA/WAM closed-loop training in procedurally generated voxel worlds.
An optimized 3D A*/Dijkstra path planner for UAVs. Features voxel-based grid search, 18-direction movement support, and heuristic-driven optimization for finding the mathematically shortest path in complex 3D obstacle fields.
Deterministic Q16.16 C++ kernel for drone swarm vision. Replaces stochastic AI guesswork with the Architecture of Necessity—mapping physical boundaries through spatial Laplacians and high-mass mechanical simulation. Zero-drift, bit-exact sensor fusion for industrial permanence.
An Augmented Grey Wolf Optimizer (AGWO) implementation for 3D path planning. This bio-inspired metaheuristic optimizes trajectories for UAVs and robotics in complex obstacle environments with adaptive recovery logic.
A project focused on constructing autonomous UAV navigation datasets for natural environments, leveraging generative AI to enhance and generate training data from images of Taiwan's roads and rivers.
Research and evaluation platform for stereo MSCKF visual-inertial odometry in GPS-denied UAV navigation, extended with estimator diagnostics, EuRoC benchmarking, trajectory analysis, and real-time visualization.
Python simulation studying how perception errors (false positives, false negatives, sensor noise, latency, dropout) and sensor fusion modes (no fusion, naive, trust-weighted) affect multi-UAV swarm safety and mission success. Includes a matplotlib-based visualizer for live playback and CSV replay of simulation runs.
Add a description, image, and links to the uav-navigation topic page so that developers can more easily learn about it.
To associate your repository with the uav-navigation topic, visit your repo's landing page and select "manage topics."