data-association

Structure from Motion (SfM)

B. Sc. Thesis concerning the semi-automatic creation of a digital twin of a traffic scene near Amirkabir University of Technology

Code for Learning Data Association for Multi-Object Tracking using Only Coordinates [Pattern Recognition]

State Estimation using Kalman Filters, EKF and solving the Data Association problem

Multi-sensor multi-target tracking system for harbour surveillance. Fuses mm-wave radar, stereo camera, AIS and GNSS via an Extended Kalman Filter with Hungarian (GNN) data association and M-of-N track lifecycle. Validated on simulated scenarios and real Copenhagen harbour recordings.

Modification of KITTI-360 dataset for Scene Recognition (Data Assosiation) tasks.

VIMMJIPDA multi-target tracker integrated for use with the colav-simulator. Open sourced in November 2025.

Core implementation of ClarityTrack for online multi-object tracking via confidence-aware data association.

A ROBUST MULTI-CUE HYBRID FRAMEWORK FOR MULTI-OBJECT TRACKING

Intelligence, Surveillance, Reconnaissance — sensor fusion, multi-INT fusion, track management, Kalman/EKF filters.

A script used to generate the spectral correspondence figure in my DPhil thesis.

EKF localisation and EKF-SLAM on procedural Formula-Student-style cone tracks, with NEES / NIS consistency tests over seeded Monte-Carlo runs.

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.

RGB-D perception core for constrained fruit-stem association, robust 3D localization and tracking in robotic harvesting.

Real-time causal BEV object tracking using Kalman filtering and data association for stable autonomous-vehicle perception and longitudinal control.

Multi-target tracking from the filter up: IMM estimator, JPDA data association, sequential track scoring, and OSPA evaluation that reports where each approach loses

EW-Resilient Autonomous Threat Estimation & Multi-Sensor Fusion Engine using Cross-Attention Transformers, PINN & JSBSim 6-DoF Dynamics.