fastslam

solution of exercises of the book "probabilistic robotics"

WiFi based Radio-Inertial Localization

This project implements grid-based FastSLAM1.0 and FastSLAM2.0 algorithms to solve SLAM problem in a simulated environment.

EKF-SLAM and FastSLAM1.0 implementation on the UTIAS Multi-Robot Cooperative Localization and Mapping Dataset. It also provides metrics for evaluation like: ATE, RPE and Landmark RMSE.

FastSLAM with Pioneer robot in CoppeliaSim(V-REP)

Landmark based FastSLAM implementation followed by global path planning for a small four wheeled robot

SLAM programms

SLAM Course by Cyrill Stachniss, University of Freiburg. Winter 2013. Assigments

Final project of the Udacity's Robotics Software Engineer Nanodegree.

📍 Educational FastSLAM 1.0 implementation for 2D robot localization and landmark mapping using a particle filter and Extended Kalman Filters.

Minimal SLAM framework comparing EKF-SLAM and FastSLAM 1.0 from scratch in Python. Features unknown data association, Joseph-form covariance updates, and performance benchmarking.

Map My World (project 4 of 5 from Udacity Robotics Software Engineer Nanodegree)

Motion models, a discrete Bayes filter, Monte Carlo localization, both EKF Jacobians, landmark FastSLAM and six path planners compared — with per-stage provenance

FastSLAM with per-particle landmark EKFs and selective resampling for simultaneous localization and mapping

Experiments on sampling in particle filters: how resampling schemes, proposal distributions and MCMC moves affect accuracy, particle degeneracy and runtime in robot localization and FastSLAM. Built on open-source filters, with plug-in techniques and reproducible results.