ATKS

The research focuses on applied machine intelligence technologies. Based on fundamental research, new technical systems such as autonomous vehicles, autonomous service robots or assistance systems with cognitive capabilities are to be realized. The application of these so-called technical-cognitive systems primarily takes place in the context of highly automated, efficient and intermodal mobility, networked automated production and logistics as well as the interactive support of the user in everyday situations.

The addressed foundations of machine intelligence are primarily machine perception, situational understanding and behavioral decision-making. Methods of machine learning and probabilistic inference are researched and applied for all components. System evaluation and validation methods are an additional focus of applied research. Autonomous vehicles such as CoCar NextGen, CoCar, and CoCar-Zero, autonomous shuttles such as Anna and Ella, mobile robots such as the assistance robot Hollie, the Lauron walking machines or the inspection robot Kairo form valuable integration and evaluation platforms for applied research. They are being further developed in close cooperation with the FZI and used for joint research and teaching.

Textüberschrift

Offene Abschlussarbeiten und HiWi-Stellen

Forschungsbereich	Themen
Perception	Traffic Light Detection with Deep Neural Networks for Autonomous Driving Aerial Image Segmentation with Deep Neural Networks for Autonomous Driving Deep Learning based 3D Environment Perception for Autonomous Vehicles Sensor-setup agnostic Machine Learning Perception for Autonomous Vehicles Robust, Interpretable and Energy-Efficient Deep Learning for Camera-based Perception Sensorfusion for Connected Autonomous Driving Intelligent Roadside Infrastructure for Connected Autonomous Driving Recognition of Sensor Data Discrepancies in Autonomous Vehicles and Smart Infrastructure Perception with Intelligent Traffic Infrastructure Object Detection with Uncertainty to Tackle Shifted and Out-Of-Distribution Data
Prediction	Predicting the Behavior of Traffic Participants with Deep Neural Networks Behavior and Motion Prediction of Traffic Participants for Safe Trajectory Planning Probabilistic Decision Making and Scene Interpretation A priori VRU Behavior Prediction using Traffic Infrastructure for Autonomous Driving Trajectory forecasting under road topology and vehicle constraints
UX	Automated Capture of User Experience in Autonomous Vehicles
Maps	Automated Generation and Maintenance of Probabilistic HD-Maps Aerial Image Segmentation with Deep Neural Networks for Autonomous Driving Semantic LiDAR-Localization and Validation through GPS and Odometrie
Planning	Cooperative Trajectory Planning under Uncertainties Probabilistic Decision Making and Scene Interpretation
Safety and Security	Misbehavior Detection and Optimization of Driving Components for Robustness Improvement Evaluating Metrics for Performance Assessment in Autonomous Driving Adversarial Attacks on Deep Learning Models Condition Monitoring for Robust and Safe Autonomous Systems
Vehicle-to-Everything (V2X/Car2X)	Vehicle-to-everything (V2X) for Distributed Simulations on Proving Grounds and Test Areas for Autonomous Driving Interactive Machine Learning for Remote Assisted Autonomous Vehicles Remote Operation in Autonomous Driving
Simulation	Graph-Level Anomaly Detection in Traffic Scenarios for Autonomous Driving (AD) Distributed Virtual Reality (VR) and Simulation Frameworks for Validation and Verification of Autonomous Vehicles Simulation of Autonomous Driving and Behaviour Modelling of Vulnerable Road Users
Reinforcement Learning	Distributional and Model-Based Reinforcement Learning Efficient Unsupervised Reinforcement Learning End-to-End autonomous driving using Reinforcement Learning Generally Capable Autonomous Agents from Web-Scale Video Data
Other Topics in Autonomous Driving	Postprocessing of Trajectory Tracking Vehicle Hardware and Sensor Setups for Autonomous Vehicles

ATKS

Textüberschrift

Offene Abschlussarbeiten und HiWi-Stellen

Perception

Prediction

UX

Maps

Planning

Safety and Security

Vehicle-to-Everything (V2X/Car2X)

Simulation

Reinforcement Learning

Other Topics in Autonomous Driving