Ongoing Theses

We are always searching for good students and have many good opportunities. However, we are also open to suggestions by students as long as they touch the topics relevant to intelligent systems. > Click here for Offered Topics / Angebotene Themen < Note that these Theses are only for TU Darmstadt students who can directly contact Team Members for more information. For exceptional external students, we offer Theses as well, these need to contact Jan Peters by jan.peters@tu-darmstadt.de first. Here is the list of currently ongoing theses:

StudentAdvisorTypeTopic
Marcel MittenbühlerAhmed Hendawy, Carlo D'Eramo, Georgia ChalvatzakiM.Sc.Lifelong Robot Learning with Pretrained Multimodal Models
Joshua JohannsonOleg ArenzM.Sc.Bipedal Locomotion on Bolt
Paolo MaglianoDavide Tateo from IAS, Raffaello Camoriano from PoliToM.Sc.SafeLoco: Safe Reinforcement Learning for Robotic Locomotion
Marius ZöllerKay Hansel and An Thai Le from IASM.Sc.Enhancing Smoothness in Policy Blending with Gaussian Processes
Yi LiAlexandre Chapin, Liming Chen, Emmanuel Dellandrea from LIRIS, Tim Schneider, Alap KshirsagarM.Sc.Self-supervised learning of a visual object-centric representation for robotic manipulation
Jiahui ShiJunning HuangM.Sc.Hybrid Force-Impedance Control for Robot Manipulators
Haris CavkicSuman Pal from IAS, Aiswarya Menon from IASM.Sc.Flow Matching for Industrial Motion Planning: Learning Trajectory Distributions for Real-World Applications
Vojin VlačićCristiana Mirandade Farias, Alap Kshirsagar, Tim SchneiderM.Sc.Active Exploration for Bimanual Grasping
Khrystyna RudAlap Kshirsagar, Oliver HahnM.Sc.Self-supervised Learning for Robotic Visuomotor Skills
Jonathan LippertKai Ploeger, Alap Kshirsagar, Abir ChowdhuryM.Sc.Human-Robot Partner Juggling
Tim FaustTheo VincentM.Sc.Eau De Soft Actor-Critic: Adaptive Distillation of Neural Networks in Deep Reinforcement Learning
Christian SchererDaniel Palenicek, Theo Gruner, Joe Watson from University of OxfordB.Sc.Coherent Soft Imitation Learning for Vision-Language Action Models
Henrik MetternichAhmed HendawyM.Sc.Value-based Reinforcement Learning via Mode Connectivity
Magnus DierkingAn Thai Le, Joao CarvalhoM.Sc.Model Tensor Planning
Keagan HolmesAhmed HendawyM.Sc.Sharing Multimodal Behaviors in Multi-Task Reinforcement Learning through Diffusion Models
Yiyang WangSiwei Ju from IASM.Sc.Benchmarking Reinforcement Learning and Control Methods on F1Tenth Car Model
Michelle SaiaAryaman Reddi from LiteRL/IAS, Nico Bohlinger from IASM.Sc.Targetted simulator state resetting for sample-efficient reinforcement learning
Zhijingshui YangJunning Huang from IAS, Jiayun Li from PEARLM.Sc.Hybrid Force–Impedance Control for Robotic Household Manipulation
Mustafa KamalMichael Drolet from IASM.Sc.Learning-based Methods for Gait Cycle Phase Prediction
Yusuf Bera DemirhanClaudius KienleM.Sc.Determining Optimal Viewpoints for Robust Segmentation of 3D Models
Yogesh TripathiTheo VincentM.Sc.Off-policy Reinforcement Learning: following the value iteration path
Kevin GerhardtTheo VincentB.Sc.Sample-efficient off-policy reinforcement learning with residual gradients
Matteo PrincisghDavide Tateo, Joao Carvalho from IAS, Matteo Lupperto from Università degli Studi di MilanoM.Sc.Learning to Plan

Ongoing IP Projects

TopicStudentsAdvisor
Robotic Knot UntanglingTim Missal, Tom Roehner, Gina Leonie WigginghausBerk Gueler from IAS
Long Horizon Shared AutonomyZhihao Han, Yuchao Wu, Ahmad HamadBerk Gueler from IAS
Controlling Humanoids by Natural LanguageRuixuan Ge, Khaled SadeqClaudius Kienle, Oleg Arenz
Motion Planning on Constraint ManifoldsMatej Zoltan Suvert, Philipp Schönfeld, Hangjian QianCristiana Mirandade Farias, Kay Pompetzki
Hierarchical Semantic Segmentation of Human Demonstration DataDaniel Neykov, David Lang , David Stebich, Paul KnaußtCristiana Mirandade Farias, Kay Pompetzki
Investigating neural network updates in Off-Policy Deep Reinforcement LearningHabib MaraqtenTheo Vincent
Large-scale Procedural Robot GenerationNurhak Yalcin, Lukas MüllerNico Bohlinger