Daniel Palenicek

A robot operating system framework for using large language models in embodied AI

Christopher E. Mower, Yuhui Wan, Hongzhan Yu, Antoine Grosnit et autres

gb, de, ch (code pays fourni par la source)

StructSAM: Structure- and Spectrum-Preserving Token Merging for Segment Anything Models

Duy M. H. Nguyen, Tuan A. Tran, Duong Nguyen, Siwei Xie et autres

Recent token merging techniques for Vision Transformers (ViTs) provide substantial speedups by reducing the number of tokens processed by self-attention, often without retraining. However, their direct application to the Segment Anything Model (SAM) family is nontrivial: SAM's image encoder mixes windowed and …

StructSAM: Structure- and Spectrum-Preserving Token Merging for Segment Anything Models

Duy M. H. Nguyen, Tuan A. Tran, Duong Nguyen, Siwei Xie et autres

Recent token merging techniques for Vision Transformers (ViTs) provide substantial speedups by reducing the number of tokens processed by self-attention, often without retraining. However, their direct application to the Segment Anything Model (SAM) family is nontrivial: SAM's image encoder mixes windowed and …

de, au, us, vn, ru, hu (code pays fourni par la source)

Diminishing Return of Value Expansion Methods

Daniel Palenicek, Michael Lutter, J. Carvalho, Daniel Dennert et autres

Model-based reinforcement learning aims to increase sample efficiency, but the accuracy of dynamics models and the resulting compounding errors are often seen as key limitations. This paper empirically investigates potential sample efficiency gains from improved dynamics models in model-based value expansion methods. …

de (code pays fourni par la source)

Gait in Eight: Efficient On-Robot Learning for Omnidirectional Quadruped Locomotion

Nico Bohlinger, Jonathan Kinzel, Daniel Palenicek, Łukasz Antczak et autres

On-robot Reinforcement Learning is a promising approach to train embodiment-aware policies for legged robots. However, the computational constraints of real-time learning on robots pose a significant challenge. We present a framework for efficiently learning quadruped locomotion in just 8 minutes of raw …

de, pl (code pays fourni par la source)

XQC: Well-conditioned Optimization Accelerates Deep Reinforcement Learning

Daniel Palenicek, Florian Vogt, Joe Watson, Ingmar Posner et autres

Sample efficiency is a central property of effective deep reinforcement learning algorithms. Recent work has improved this through added complexity, such as larger models, exotic network architectures, and more complex algorithms, which are typically motivated purely by empirical performance. We take a …

Scaling CrossQ with Weight Normalization

Daniel Palenicek, Florian Vogt, Jan Peters

Reinforcement learning has achieved significant milestones, but sample efficiency remains a bottleneck for real-world applications. Recently, CrossQ has demonstrated state-of-the-art sample efficiency with a low update-to-data (UTD) ratio of 1. In this work, we explore CrossQ's scaling behavior with higher UTD ratios. …

Towards Safe Robot Foundation Models Using Inductive Biases

Maximilian Tölle, Theo Gruner, Daniel Palenicek, Tim Schneider et autres

Safety is a critical requirement for the real-world deployment of robotic systems. Unfortunately, while current robot foundation models show promising generalization capabilities across a wide variety of tasks, they fail to address safety, an important aspect for ensuring long-term operation. Current robot …

ROS-LLM: A Framework for Embodied AI

Christopher E. Mower, Yuhui Wan, Hongzhan Yu, Antoine Grosnit et autres

se, gb, de, ch, es (code pays fourni par la source)

Gait in Eight: Efficient On-Robot Learning for Omnidirectional Quadruped Locomotion

Nico Bohlinger, Jonathan Kinzel, Daniel Palenicek, Łukasz Antczak et autres

On-robot Reinforcement Learning is a promising approach to train embodiment-aware policies for legged robots. However, the computational constraints of real-time learning on robots pose a significant challenge. We present a framework for efficiently learning quadruped locomotion in just 8 minutes of raw …

Scaling Off-Policy Reinforcement Learning with Batch and Weight Normalization

Daniel Palenicek, Florian Vogt

Reinforcement learning has achieved significant milestones, but sample efficiency remains a bottleneck for real-world applications. Recently, CrossQ has demonstrated state-of-the-art sample efficiency with a low update-to-data (UTD) ratio of 1. In this work, we explore CrossQ's scaling behavior with higher UTD ratios. …

us, se, de (code pays fourni par la source)

Analysing the Interplay of Vision and Touch for Dexterous Insertion Tasks

Janis Lenz, Theo Gruner, Daniel Palenicek, Tim Schneider et autres

Robotic insertion tasks remain challenging due to uncertainties in perception and the need for precise control, particularly in unstructured environments. While humans seamlessly combine vision and touch for such tasks, effectively integrating these modalities in robotic systems is still an open problem. …