Open-source algorithms and resources
List of open-source algorithms and resources for autonomous drones. The list is a work in progress!
Perception
Link | Who | Description | ROS |
---|---|---|---|
Great Roadmap for Visual SLAM | |||
ETH | voxel-based mapping | ✔️ | |
ETH | visual inertial mapping | ✔️ | |
sparse 3D reconstruction | ✔️ | ||
U. of Delaware | EKF fuses inertial info with sparse visual features | ✔️ | |
ETH | semi-direct paradigm to estimate pose from pixel intensities and features | ✔️ | |
TUM | direct sparse odometry | ||
UCLA | inertial-aided visual odometry | ||
HKUST | An optimization-based multi-sensor state estimator | ||
MIT | real-time metric-semantic SLAM and VIO | ✔️ | |
UPenn | tagSLAM with apriltags | ✔️ | |
A lightweight, accurate and robust monocular visual inertial odometry based on Multi-State Constraint Kalman Filter. | ✔️ | ||
based on robocentric sliding-window filtering-based VIO framework | |||
MIT | fast, uncertainty-aware proximity queries with lazy search of local 3D data | ||
UPenn | package is a stereo version of MSCKF | ||
HKUST | Robust and Versatile Monocular Visual-Inertial State Estimator | ||
Simbe Robotics / Samsung Research | SLAM for massive maps | ✔️ |
Navigation
Link | Who | Description | ROS |
---|---|---|---|
ETH | creates polynomial path | ✔️ | |
ETH | planning tool using voxblox (RRT*, etc.) | ✔️ | |
ETH | deep learning visual navigation | ||
TUM | replanning of global traj, needs prior map | ||
Georgia Tech | end-to-end navigation trained from simulation | ||
Autonomous navigation for drones | |||
HKUST | robust and efficient trajectory planner for quads | ✔️ | |
Zhejiang University | Autonomous and Decentralized Quadrotor Swarm System in Cluttered Environments | ||
HKUST | Safe Trajectory Generation For Complex Urban Environments Using Spatio-temporal Semantic Corridor | ||
ETH | obstacle avoidance with event cameras | ||
KTH | unknown environment exploration based on octomap | ||
ETH | unknown environment exploration | ||
HKUST | a complete and robust system for aggressive flight in complex environment | ||
ETH | deep learning Sim2Real Drone racing | ✔️ | |
ETH | high-level waypoint-following for micro aerial vehicles | ✔️ | |
Eureka Robotics | Time-Optimal Path Parameterization |
Controls
Link | Who | Description | ROS |
---|---|---|---|
Berkeley | Model Predictive Control with one-step feedforward neural network dynamics model from Model-based Reinforcement Learning | ||
ETH | efficient C++ library for control, estimation, optimization and motion planning in robotics | ||
library implementing the linear and nonlinear control theories in python | |||
ETH | Model Predictive Control for Quadrotors with extension to Perception-Aware MPC | ||
ETH | alternative to PX4 that works with RotorS | ||
flight control tuning framework with a focus in attitude control | |||
MPC toolkit that takes care of the implementation | |||
ETH | also has PX4 implementation (claim badly hacked though) | ||
ETH | Data-Driven MPC for Quadrotors | ||
ETH | fly complex maneuvers with multi-layer perceptron | ||
ETH | trajectory tracking with MPC | ||
ETH | complete framework for flying quadrotors | ||
HKUST | high level controller compatible with DJI N3 flight controller | ||
combines mav_trajectory_generation and waypoint_navigator with mavros_controller | ✔️ | ||
ETH | calculates model parameters for a drone | ||
CTU | framework for controlling drones with PX4 and different advanced controllers |
Labs/Organizations to follow
Lab Website | Git | Where |
---|---|---|
Zurich, Switzerland | ||
Philadelphia, USA | ||
Hangzhou, China | ||
Clear Water Bay, Hong Kong | ||
Kanpur, India |
Last updated