Learning human-to-robot handovers through 3D scene reconstruction

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.

Bio-inspired soft robotic swimmers offer a promising route toward efficient and agile underwater locomotion. Yet, most existing designs rely on single degree of freedom tails that couple propulsion and steering, limiting controllability and efficiency. In this work, we present a compliant, tendondriven robotic fish featuring an independently actuated caudal fin that decouples tail bending and fin orientation. Actuated via fishing cables, all motors are housed within the central body, minimizing rear-end ine

Estimating 3D human poses from 2D images remains challenging due to occlusions and projective ambiguity. Multi-view learning-based approaches mitigate these issues but often fail to generalize to real-world scenarios, as large-scale multi-view datasets with 3D ground truth are scarce and captured under constrained conditions. To overcome this limitation, recent methods rely on 2D pose estimation combined with 2D-to-3D pose lifting trained on synthetic data. Building on our previous MPL framework, we propose

Social robots offer promising possibilities for special education, particularly in supporting children with complex neurodevelopmental disorders (NDDs) who often experience learning difficulties, such as challenges with handwriting skills. However, most studies to date have been short-term and conducted in controlled lab settings. This study presents a robot-assisted intervention using a Wizard of Oz interface, the R2C3 system (Rehabilitation Robotic Companion for Children and Caregivers), to support handwr

Classical aerial robots have demonstrated precise control and high speeds through the use of highly optimized rigid body structures. However, when compared to their biological flying counterparts, they still struggle to match performance parameters in energy efficiency, multi-functionality, multi-terrain accessibility, scalability (to micro-scale), agility, close contact interaction, robustness, and maneuverability in cluttered environments. The recent emergence of soft aerial robotics has seen an increased

Slow-speed animals can also exhibit remarkable capabilities, as seen in snails that crawl while maintaining adhesion. Snails have inspired researchers to develop traveling wave-based robots and suction robots; however, the combination of traveling wave propulsion with suction ability remains a challenge. In this paper, we propose a snail-inspired robot that integrates a corkscrew propulsion mechanism with distributed suction cups, enabling it to crawl upside down on the ceiling. The propulsion model of the

Next-generation motion control systems require fast and precise control. However, advanced control strategies often rely on complex and costly system models.Data-driven methods have been proposed to design high-performance controllers without requiring a parametric model of the system. In particular, methods using frequency response functions (FRFs) have been widely applied to mechatronic systems due to their good performance, and the industry's familiarity with obtaining FRFs.This paper applies a recently

Trajectory optimization for motion planning requires good initial guesses to obtain good performance. In our proposed approach, we build a memory of motion based on a database of robot paths to provide good initial guesses. The memory of motion relies on function approximators and dimensionality reduction techniques to learn the mapping between the tasks and the robot paths. Three function approximators are compared: k-Nearest Neighbor, Gaussian Process Regression, and Bayesian Gaussian Mixture Regression.

Identifying the communication channels inside a society of zebrafish could allow scientists to trigger specific behaviors on the animals to tests for instance the effect of a drug or a genetic mutation. We propose a robotic system to integrate the zebrafish shoal in a non-invasive way, in order to be accepted by the animals and influence their behavior. This system is composed of two parts: a mobile robot that can reach speed of 50 cms-1 and acceleration of 2 m/s2 which are typical for the zebrafish and a l

Characterization of gait is largely used for research on locomotion or for clinical improvement and follow up. There exists two main technologies for measurement of the kinematics of walking: IMU systems and optical tracking. Inertial measurement units are getting more and more popular because of their integrability, cost and lack of spatial limitations. Despite these advantages, optical motion tracking solutions offer a higher precision which is for various applications such as kinetics analysis. In that s

Face rendering is a really important topic in Computer Graphics because a lot of virtual simulations or video games contain virtual humans. In order to obtain a realistic face, we need to take care of skin rendering. Nowadays, we can use modern 3D scanning technology to obtain very detailed meshes and textures for the face but the main difficulty with skin rendering is that we need to model subsurface scattering effects. In 2007, d'Eon, Eugene, David Luebke, and Eric Enderton published the article "Efficien

In this abstract, we combine work from [Lagarde et al., 2010] and [Calinon et al., 2009] for learning and reproduction of, respectively, navigation tasks on a mobile robot and gestures with a robot arm. Both approaches build a sensory motor map under human guidance to learn the desired behavior. With several actions possible at the same time, the selec- tion of action becomes a real issue. Several solutions exist to this problem : hi- erarchical architecture, parallel modules including subsumption architect

Localization and context interpretation are two key competences for mobile robot systems. Visual place recognition, as opposed to purely geometrical models, holds promise of higher flexibility and association of semantics to the model. Ideally, a place recognition algorithm should be robust to dynamic changes and it should perform consistently when recognizing a room (for instance a corridor) in different geographical locations. Also, it should be able to categorize places, a crucial capability for transfer

The manipulation and assembly of microsystems (MEMS) require accurate robots operating in a clean room environment. Available systems are bulky, thus requiring large ground surface for small components. Furthermore they have limited accuracy. The main objective of the 'Pocket-factory' project is to develop miniaturized and modular clean production units for microsystems and establish a methodology to quickly set them up

This paper presents general ideas on design and implementation of robots to be used together with groups of animals. This work is part of the European project LEURRE aiming to study, model and control mixed societies of animals and robots. After a short presentation of the general project, the implications on the special requirements will be discussed. The design process of the robot will be described leading to the implementation of the autonomous mini robot called InsBot. This is very compact (19x30x41 mm

To fulfill a need for natural, user-friendly means of interacting and reprogramming toy and humanoid robots, a growing trend of robotics research investigates the integration of methods for gesture recognition and natural speech processing. Unfortunately, efficient methods for speech and vision processing remain computationally expensive and, thus, cannot be easily exploited on cost- and size-limited platforms. Personal Digital Assistants (PDAs) are ideal low-cost platforms to provide simple speech and visi

With recent progress in computing, algorithmics and telecommunications, 3D models are increasingly used in various multimedia applications. Examples include visualization, gaming, entertainment and virtual reality. In the multimedia domain 3D models have been traditionally represented as polygonal meshes. This piecewise planar representation can be thought of as the analogy of bitmap images for 3D surfaces. As bitmap images, they enjoy great flexibility and are particularly well suited to describing informa

In this paper we present an autonomous mobile robot system allowing for complex collaborative interaction with non-experienced users while giving a pre-programmed tour of a public exposition. For this several modalities of man-machine communication are used simultaneously. To ensure secure and reliable operation we present a layered development approach centered on the programming system SOUL. At the Swiss National Exhibition Expo.02, ten RoboX systems will interact with hundreds of visitors per day, during

The Autonomous Systems Lab at the Swiss Federal Institute of Technology Lausanne (EPFL) is engaged in mobile robotics research. The lab s research focuses mainly on indoor localization and map building, outdoor locomotion and navigation, and micro mobile robotics. In the framework of a research project on mobile robot localization, a graphical web interface for our indoor robots has been developed. The purpose of this interface is twofold: it serves as a tool for task supervision for the researcher, and for

We propose an algorithm for aligning a preoperative computed tomography (CT) volume and intraoperative C-arm images, with applications in computer-assisted spinal surgery. Our three-dimensional (3D)/two-dimensional (2D) registration algorithm is based on splines and is tuned to a multiresolution strategy. Its goal is to establish the mutual relations of locations in the real-world scene to locations in the 3D CT and inthe 2D C-arm images. The principle of the solution is to simulate a series of C-arm images

This paper is concerned with adaptation capabilities of evolved neural controllers. A method consisting of encoding a set of local adaptation rules that synapses obey while the robot freely moves in the environment [6] is compared to a standard fixed-weight network. In the experiments presented here, the performance of the robot is measured in environments that are different in significant ways from those used during evolution. The results show that evolutionary adaptive controllers can adapt to environment

Learning human-to-robot handovers through 3D scene reconstruction

Chat with Graph Search

Improving Efficiency and Maneuverability through Design of a Robotic Fish with Independent Control of Tail and Caudal Fin Deflection

RUMPL: Ray-Based Transformers for Universal Multi-View 2D to 3D Human Pose Lifting

Robot-assisted handwriting training: An intervention for children with neurodevelopmental disorders

Soft Aerial Robots: The Development of Adaptive, Multi-Functional and Multi-Terrain Aerial Robots

Observation of Snails and a Bionic Snail Robot Crawling with Distributed Suction

High Precision Robotic Arm Control Using Frequency-Based Data-Driven Methods

Memory of Motion for Warm-Starting Trajectory Optimization

Interacting with zebrafish using robotic agents

A robotic platform for lower limb optical motion tracking in open space

Photo-Realistic Real-time Face Rendering

A state-action neural network supervising navigation and manipulation behaviors for complex task reproduction

Towards Robust Place Recognition for Robot Localization

“Pocket Factory”: Concept of miniaturized modular cleanrooms

InsBot: Design of an Autonomous Mini Mobile Robot Able to Interact with Cockroaches

PDA Interface for Humanoid Robots

Compression of 3D models with NURBS

The interactive autonomous mobile system RoboX

Web Interfacing for Task Supervision and Specification

Multiresolution Spline-Based 3D/2D Registration of CT Volume and C-Arm Images for Computer-Assisted Surgery

Evolutionary Robots with Fast Adaptive Behavior in New Environments