Golnaz Raja

"Doctoral researcher in machine learning, robotics, and safe autonomous navigation."

Ghost

Uncertainty-aware autonomous navigation

I am a doctoral researcher in Machine Learning and Robotics at Tampere University. My work combines probabilistic machine learning, perception, mapping, and risk-aware planning to make autonomous robots safer and more reliable on challenging rough terrain.

Experiences

Tampere University

Doctoral Researcher

August 2024 - Present
Researching uncertainty-aware safe autonomous navigation on rough terrain. I design probabilistic traversability maps and risk-aware planners as part of the Finnish Doctoral Program Network in AI (AI-DOC).
Probabilistic ML
Path Planning
Robotics
PyTorch
ROS 2
Tampere University

Researcher

December 2023 - August 2024
Designed a sensor-agnostic, perception-based navigation system and developed rough-terrain models in CARLA Simulator and Unreal Engine.
CARLA
Unreal Engine
Autonomous Navigation
Perception
Tampere University

Research Assistant

November 2021 - December 2023
Developed vision-based obstacle avoidance with Control Barrier Functions, optimization-based robot controllers, semantic segmentation, and learning-based navigation systems.
CBF
Computer Vision
CARLA
Unity
RGB-D
GANs
Tampere University

Master's Thesis Worker

July 2023 - October 2023
Designed vision-based cost maps and a scenario-based evaluation system for a V-CBF safe controller, then deployed the controller on an industrial mobile robot.
V-CBF
CARLA
Computer Vision
Mobile Robotics
ARAS Group

Researcher

October 2019 - September 2021
Developed a mixed-reality eye-surgery training simulation for the ARASH ASiST robot and an evaluation system based on reinforcement learning.
Mixed Reality
Unity
SOFA
Reinforcement Learning
K. N. Toosi University

Bachelor's Thesis Worker

March 2019 - May 2019
Implemented marker-based and image-based augmented-reality software using computer vision.
Augmented Reality
Computer Vision
INIT
E

Simulation Twin 1, 2 & 3

Realistic simulation and standards-based evaluation for mobile robots

Researched CARLA and AirSim, developed scenario-based autonomous-navigation evaluations based on ISO standards, designed custom CARLA maps, and integrated RoadRunner, ScenarioRunner, and OpenSCENARIO workflows.
_ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _  _ CARLA _ _ AirSim _ _ RoadRunner _ _ OpenSCENARIO _