Aleksandar Taranovic

Aleksandar Taranovic

PhD Student at ALR@KIT

Posts by Collection

portfolio

publications

Adaptable Multimodal Interaction Framework for Robot-Assisted Cognitive Training

Published in HRI 2018 (Companion), 2018

A framework that adjusts robot assistance level and action execution modalities for cognitive training of patients with mild cognitive impairment or Alzheimer.

Recommended citation: Aleksandar Taranovic, Aleksandar Jevtic, Carme Torras. (2018). "Adaptable Multimodal Interaction Framework for Robot-Assisted Cognitive Training." Companion of the 2018 ACM/IEEE International Conference on Human-Robot Interaction.
Download Paper

Resource-Based Modality Selection in Robot-Assisted Cognitive Training

Published in HRI 2018 (Companion), 2018

A resource-based approach for selecting interaction modalities in robot-assisted cognitive training scenarios.

Recommended citation: Aleksandar Taranovic, Aleksandar Jevtic, Joan Hernandez-Farigola, Natalia Tantinya, Carla Abdelnour, Carme Torras. (2018). "Resource-Based Modality Selection in Robot-Assisted Cognitive Training." Companion of the 2018 ACM/IEEE International Conference on Human-Robot Interaction.
Download Paper

Adaptive Modality Selection Algorithm in Robot-Assisted Cognitive Training

Published in IROS 2018, 2018

An algorithm that selects robot interaction modalities based on the state of the user and the environment, as well as user preferences.

Recommended citation: Aleksandar Taranovic, Aleksandar Jevtic, Carme Torras. (2018). "Adaptive Modality Selection Algorithm in Robot-Assisted Cognitive Training." IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).
Download Paper

Automatic Learning of Cognitive Exercises for Socially Assistive Robotics

Published in RO-MAN 2021, 2021

Automatic generation and adaptation of cognitive exercises for socially assistive robots.

Recommended citation: Alejandro Suarez-Hernandez, Antonio Andriella, Aleksandar Taranovic, Javier Segovia-Aguas, Carme Torras, Guillem Alenya. (2021). "Automatic Learning of Cognitive Exercises for Socially Assistive Robotics." IEEE International Conference on Robot and Human Interactive Communication (RO-MAN).
Download Paper

Adversarial Imitation Learning with Preferences

Published in ICLR 2023, 2023

Extending adversarial imitation learning to simultaneously utilize both demonstrations and preferences for policy learning.

Recommended citation: Aleksandar Taranovic, Andras Gabor Kupcsik, Niklas Freymuth, Gerhard Neumann. (2023). "Adversarial Imitation Learning with Preferences." International Conference on Learning Representations (ICLR).
Download Paper

Towards Fusing Point Cloud and Visual Representations for Imitation Learning

Published in arXiv preprint, 2025

Combining point cloud and image representations to get the best of both modalities for imitation learning.

Recommended citation: Atalay Donat, Xiaogang Jia, Xi Huang, Aleksandar Taranovic, Denis Blessing, Ge Li, Hongyi Zhou, Hanyi Zhang, Rudolf Lioutikov, Gerhard Neumann. (2025). "Towards Fusing Point Cloud and Visual Representations for Imitation Learning." arXiv preprint arXiv:2502.12320.
Download Paper

AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction

Published in NeurIPS 2025, 2025

A supervised learning approach to adaptive mesh generation that iteratively predicts sizing fields from expert demonstrations.

Recommended citation: Niklas Freymuth, Tobias Wuerth, Nicolas Schreiber, Balazs Gyenes, Andreas Boltres, Johannes Mitsch, Aleksandar Taranovic, Tai Hoang, Philipp Dahlinger, Philipp Becker, Luise Kaerger, Gerhard Neumann. (2025). "AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction." Conference on Neural Information Processing Systems (NeurIPS).
Download Paper

Scaffolding Dexterous Manipulation with Vision-Language Models

Published in NeurIPS 2025, 2025

Using vision-language models to specify tasks and guide exploration for dexterous robotic manipulation without human demonstrations or handcrafted rewards.

Recommended citation: Vincent de Bakker, Joey Hejna, Tyler Ga Wei Lum, Onur Celik, Aleksandar Taranovic, Denis Blessing, Gerhard Neumann, Jeannette Bohg, Dorsa Sadigh. (2025). "Scaffolding Dexterous Manipulation with Vision-Language Models." Conference on Neural Information Processing Systems (NeurIPS).
Download Paper

PAWS: Preference Learning with Advantage-Weighted Segments

Published in ICML 2026, 2026

A preference-based RL method that aligns utility training with policy optimization using segment-level advantage functions.

Recommended citation: Aleksandar Taranovic, Onur Celik, Niklas Freymuth, Ge Li, Serge Thilges, Huy Le, Tai Hoang, Rania Rayyes, Gerhard Neumann. (2026). "PAWS: Preference Learning with Advantage-Weighted Segments." International Conference on Machine Learning (ICML).
Download Paper

talks

teaching