Leonel Rozo
YOU?
Author Swipe
View article: Diffeomorphic Obstacle Avoidance for Contractive Dynamical Systems via Implicit Representations
Diffeomorphic Obstacle Avoidance for Contractive Dynamical Systems via Implicit Representations Open
View article: Fast and Robust Visuomotor Riemannian Flow Matching Policy
Fast and Robust Visuomotor Riemannian Flow Matching Policy Open
Diffusion-based visuomotor policies excel at learning complex robotic tasks by effectively combining visual data with high-dimensional, multi-modal action distributions. However, diffusion models often suffer from slow inference due to cos…
View article: Fast and Robust Visuomotor Riemannian Flow Matching Policy
Fast and Robust Visuomotor Riemannian Flow Matching Policy Open
Diffusion-based visuomotor policies excel at learning complex robotic tasks by effectively combining visual data with high-dimensional, multi-modal action distributions. However, diffusion models often suffer from slow inference due to cos…
View article: Extended Neural Contractive Dynamical Systems: On Multiple Tasks and Riemannian Safety Regions
Extended Neural Contractive Dynamical Systems: On Multiple Tasks and Riemannian Safety Regions Open
Stability guarantees are crucial when ensuring that a fully autonomous robot does not take undesirable or potentially harmful actions. We recently proposed the Neural Contractive Dynamical Systems (NCDS), which is a neural network architec…
View article: On Probabilistic Pullback Metrics for Latent Hyperbolic Manifolds
On Probabilistic Pullback Metrics for Latent Hyperbolic Manifolds Open
Probabilistic Latent Variable Models (LVMs) excel at modeling complex, high-dimensional data through lower-dimensional representations. Recent advances show that equipping these latent representations with a Riemannian metric unlocks geome…
View article: The GeometricKernels Package: Heat and Matérn Kernels for Geometric Learning on Manifolds, Meshes, and Graphs
The GeometricKernels Package: Heat and Matérn Kernels for Geometric Learning on Manifolds, Meshes, and Graphs Open
Kernels are a fundamental technical primitive in machine learning. In recent years, kernel-based methods such as Gaussian processes are becoming increasingly important in applications where quantifying uncertainty is of key interest. In se…
View article: Riemannian Flow Matching Policy for Robot Motion Learning
Riemannian Flow Matching Policy for Robot Motion Learning Open
We introduce Riemannian Flow Matching Policies (RFMP), a novel model for learning and synthesizing robot visuomotor policies. RFMP leverages the efficient training and inference capabilities of flow matching methods. By design, RFMP inheri…
View article: Neural Contractive Dynamical Systems
Neural Contractive Dynamical Systems Open
Stability guarantees are crucial when ensuring a fully autonomous robot does not take undesirable or potentially harmful actions. Unfortunately, global stability guarantees are hard to provide in dynamical systems learned from data, especi…
View article: Unraveling the Single Tangent Space Fallacy: An Analysis and Clarification for Applying Riemannian Geometry in Robot Learning
Unraveling the Single Tangent Space Fallacy: An Analysis and Clarification for Applying Riemannian Geometry in Robot Learning Open
In the realm of robotics, numerous downstream robotics tasks leverage machine learning methods for processing, modeling, or synthesizing data. Often, this data comprises variables that inherently carry geometric constraints, such as the un…
View article: Reactive motion generation on learned Riemannian manifolds
Reactive motion generation on learned Riemannian manifolds Open
In recent decades, advancements in motion learning have enabled robots to acquire new skills and adapt to unseen conditions in both structured and unstructured environments. In practice, motion learning methods capture relevant patterns an…
View article: Wasserstein Gradient Flows for Optimizing Gaussian Mixture Policies
Wasserstein Gradient Flows for Optimizing Gaussian Mixture Policies Open
Robots often rely on a repertoire of previously-learned motion policies for performing tasks of diverse complexities. When facing unseen task conditions or when new task requirements arise, robots must adapt their motion policies according…
View article: The e-Bike Motor Assembly: Towards Advanced Robotic Manipulation for Flexible Manufacturing
The e-Bike Motor Assembly: Towards Advanced Robotic Manipulation for Flexible Manufacturing Open
Robotic manipulation is currently undergoing a profound paradigm shift due to the increasing needs for flexible manufacturing systems, and at the same time, because of the advances in enabling technologies such as sensing, learning, optimi…
View article: The E-Bike Motor Assembly: Towards Advanced Robotic Manipulation for Flexible Manufacturing
The E-Bike Motor Assembly: Towards Advanced Robotic Manipulation for Flexible Manufacturing Open
View article: Learning Riemannian Stable Dynamical Systems via Diffeomorphisms
Learning Riemannian Stable Dynamical Systems via Diffeomorphisms Open
Dexterous and autonomous robots should be capable of executing elaborated dynamical motions skillfully. Learning techniques may be leveraged to build models of such dynamic skills. To accomplish this, the learning model needs to encode a s…
View article: Optimizing Demonstrated Robot Manipulation Skills for Temporal Logic Constraints
Optimizing Demonstrated Robot Manipulation Skills for Temporal Logic Constraints Open
For performing robotic manipulation tasks, the core problem is determining suitable trajectories that fulfill the task requirements. Various approaches to compute such trajectories exist, being learning and optimization the main driving te…
View article: Bringing motion taxonomies to continuous domains via GPLVM on hyperbolic manifolds
Bringing motion taxonomies to continuous domains via GPLVM on hyperbolic manifolds Open
Human motion taxonomies serve as high-level hierarchical abstractions that classify how humans move and interact with their environment. They have proven useful to analyse grasps, manipulation skills, and whole-body support poses. Despite …
View article: Optimizing Demonstrated Robot Manipulation Skills for Temporal Logic Constraints
Optimizing Demonstrated Robot Manipulation Skills for Temporal Logic Constraints Open
For performing robotic manipulation tasks, the core problem is determining suitable trajectories that fulfill the task requirements. Various approaches to compute such trajectories exist, being learning and optimization the main driving te…
View article: Geometry-aware Bayesian Optimization in Robotics using Riemannian\n Mat\\'ern Kernels
Geometry-aware Bayesian Optimization in Robotics using Riemannian\n Mat\\'ern Kernels Open
Bayesian optimization is a data-efficient technique which can be used for\ncontrol parameter tuning, parametric policy adaptation, and structure design in\nrobotics. Many of these problems require optimization of functions defined on\nnon-…
View article: Orientation Probabilistic Movement Primitives on Riemannian Manifolds
Orientation Probabilistic Movement Primitives on Riemannian Manifolds Open
Learning complex robot motions necessarily demands to have models that are able to encode and retrieve full-pose trajectories when tasks are defined in operational spaces. Probabilistic movement primitives (ProMPs) stand out as a principle…
View article: Study of Signal Temporal Logic Robustness Metrics for Robotic Tasks\n Optimization
Study of Signal Temporal Logic Robustness Metrics for Robotic Tasks\n Optimization Open
Signal Temporal Logic (STL) is an efficient technique for describing temporal\nconstraints. It can play a significant role in robotic manipulation, for\nexample, to optimize the robot performance according to task-dependent metrics.\nIn th…
View article: Learning Forceful Manipulation Skills from Multi-modal Human Demonstrations
Learning Forceful Manipulation Skills from Multi-modal Human Demonstrations Open
Learning from Demonstration (LfD) provides an intuitive and fast approach to program robotic manipulators. Task parameterized representations allow easy adaptation to new scenes and online observations. However, this approach has been limi…
View article: Learning Riemannian Manifolds for Geodesic Motion Skills
Learning Riemannian Manifolds for Geodesic Motion Skills Open
For robots to work alongside humans and perform in unstructured environments, they must learn new motion skills and adapt them to unseen situations on the fly. This demands learning models that capture relevant motion patterns, while offer…
View article: Emerging Paradigms for Robotic Manipulation: From the Lab to the Productive World [From the Guest Editors]
Emerging Paradigms for Robotic Manipulation: From the Lab to the Productive World [From the Guest Editors] Open
The articles in this special section aim to stimulate and gather publications describing how new approaches in the field of robotic manipulation can be (or have already been) transferred from research labs to the productive world. Novel ro…
View article: Analysis and Transfer of Human Movement Manipulability in Industry-like Activities
Analysis and Transfer of Human Movement Manipulability in Industry-like Activities Open
Humans exhibit outstanding learning, planning and adaptation capabilities while performing different types of industrial tasks. Given some knowledge about the task requirements, humans are able to plan their limbs motion in anticipation of…
View article: Learning and Sequencing of Object-Centric Manipulation Skills for Industrial Tasks
Learning and Sequencing of Object-Centric Manipulation Skills for Industrial Tasks Open
Enabling robots to quickly learn manipulation skills is an important, yet challenging problem. Such manipulation skills should be flexible, e.g., be able adapt to the current workspace configuration. Furthermore, to accomplish complex mani…
View article: High-Dimensional Bayesian Optimization via Nested Riemannian Manifolds
High-Dimensional Bayesian Optimization via Nested Riemannian Manifolds Open
Despite the recent success of Bayesian optimization (BO) in a variety of applications where sample efficiency is imperative, its performance may be seriously compromised in settings characterized by high-dimensional parameter spaces. A sol…
View article: Geometry-aware manipulability learning, tracking, and transfer
Geometry-aware manipulability learning, tracking, and transfer Open
Body posture influences human and robot performance in manipulation tasks, as appropriate poses facilitate motion or the exertion of force along different axes. In robotics, manipulability ellipsoids arise as a powerful descriptor to analy…
View article: Learning and Sequencing of Object-Centric Manipulation Skills for\n Industrial Tasks
Learning and Sequencing of Object-Centric Manipulation Skills for\n Industrial Tasks Open
Enabling robots to quickly learn manipulation skills is an important, yet\nchallenging problem. Such manipulation skills should be flexible, e.g., be able\nadapt to the current workspace configuration. Furthermore, to accomplish\ncomplex m…
View article: Analysis and Transfer of Human Movement Manipulability in Industry-like\n Activities
Analysis and Transfer of Human Movement Manipulability in Industry-like\n Activities Open
Humans exhibit outstanding learning, planning and adaptation capabilities\nwhile performing different types of industrial tasks. Given some knowledge\nabout the task requirements, humans are able to plan their limbs motion in\nanticipation…
View article: Interactive Trajectory Adaptation through Force-guided Bayesian Optimization
Interactive Trajectory Adaptation through Force-guided Bayesian Optimization Open
Flexible manufacturing processes demand robots to easily adapt to changes in the environment and interact with humans. In such dynamic scenarios, robotic tasks may be programmed through learning-from-demonstration approaches, where a nomin…