Prediction of Human Whole-Body Movements with AE- ProMPs

Denny, Oriane; Chaveroche, Maxime; Colas, Francis; Charpillet, François; Ivaldi, Serena

doi:10.1109/humanoids.2018.8624986

Cited by 6 publications

(9 citation statements)

References 16 publications

(31 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Human models with a high number of degrees of freedom impact the time performance of motion analysis. To improve speed in the motion processing of the DHM data, recent works have used a representation of the human motion in a latent space [ 20 , 21 ]. Marin et al [ 20 ] showed that the use of a latent space human representation improved the performance of their application.…”

Section: Related Workmentioning

confidence: 99%

“…For this reason, they have been widely used for reducing the dimensionality of the human state and for movement generation [ 19 ]. Dermy et al [ 21 ] address the problem of predicting future human whole-body movements given prior observations. They map high-dimensionalality trajectories into a reduced latent space using AE.…”

Section: Related Workmentioning

confidence: 99%

“…To construct the LEM, we project a given ergonomics score onto a 2D latent space that represents the human posture. In our previous work, we proposed to use latent representations to capture the human movement/posture [ 21 ]. Here, we extend the use of latent maps to encode ergonomics evaluation scores.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Latent Ergonomics Maps: Real-Time Visualization of Estimated Ergonomics of Human Movements

Vianello

Gomes

Stulp

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

Improving the ergonomy of working environments is essential to reducing work-related musculo-skeletal disorders. We consider real-time ergonomic feedback a key technology for achieving such improvements. To this end, we present supportive tools for online evaluation and visualization of strenuous efforts and postures of a worker, also when physically interacting with a robot. A digital human model is used to estimate human kinematics and dynamics and visualize non-ergonomic joint angles, based on the on-line data acquired from a wearable motion tracking device.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Latent Ergonomics Maps: Real-Time Visualization of Estimated Ergonomics of Human Movements

Vianello

Gomes

Stulp

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…The critical aspect that distinguishes a robot motion controller for cooperation with humans from a generic one for the robot alone in the environment is to consider the human in the design of the motion control, i.e., to design a "human-aware" controller [27]. This means to consider the human state, their dynamics [28], their intended movement [29], and use the predictions of their future states to plan suitable robot motions and physical interactions. These interactions often result in complex behaviors, where the humanoid needs to simultaneously control various aspects of its internal and external motion like locomotion, posture, gaze, manipulation, and contact stability.…”

Section: Motion Control For Physical Interactionmentioning

confidence: 99%

Human-Humanoid Interaction and Cooperation: a Review

et al. 2021

Self Cite

View full text Add to dashboard Cite

Purpose of Review Humanoid robots are versatile platforms with the potential to assist humans in several domains, from education to healthcare, from entertainment to the factory of the future. To find their place into our daily life, where complex interactions and collaborations with humans are expected, their social and physical interaction skills need to be further improved. Recent FindingsThe hallmark of humanoids is their anthropomorphic shape, which facilitates the interaction but at the same time increases the expectations of the human in terms of advanced cooperation capabilities. Cooperation with humans requires an appropriate modeling and real-time estimation of the human state and intention. This information is required both at a high level by the cooperative decision-making policy and at a low level by the interaction controller that implements the physical interaction. Real-time constraints induce simplified models that limit the decision capabilities of the robot during cooperation. SummaryIn this article, we review the current achievements in the context of human-humanoid interaction and cooperation. We report on the cognitive and cooperation skills that the robot needs to help humans achieve their goals, and how these high-level skills translate into the robot's low-level control commands. Finally, we report on the applications of humanoid robots as humans' companions, co-workers, or avatars. Keywords Humanoid robots • Human-robot interaction • CooperationThis article is part of the Topical Collection on Humanoid and Bipedal Robotics

show abstract

“…In [11] the authors proposed the use of a time-dependent variational autoencoder to address the generalization challenges. Dermy et al [28] used auto-encoders with ProMPs for efficient human motion prediction. Colomé et al [6] have proposed a linear DR in the joint space for learning DMPs, however, in [29], the authors proposed a reduction of the parametric space, as more appropriate for learning MPs.…”

Section: Related Workmentioning

confidence: 99%

Contextual Latent-Movements Off-Policy Optimization for Robotic Manipulation Skills

Tosatto¹,

Chalvatzaki²,

Peters³

2020

Preprint

View full text Add to dashboard Cite

Parameterized movement primitives have been extensively used for imitation learning of robotic tasks. However, the high-dimensionality of the parameter space hinders the improvement of such primitives in the reinforcement learning (RL) setting, especially for learning with physical robots. In this paper we propose a novel view on handling the demonstrated trajectories for acquiring low-dimensional, non-linear latent dynamics, using mixtures of probabilistic principal component analyzers (MPPCA) on the movements' parameter space. Moreover, we introduce a new contextual off-policy RL algorithm, named LAtent-Movements Policy Optimization (LAMPO). LAMPO can provide gradient estimates from previous experience using self-normalized importance sampling, hence, making full use of samples collected in previous learning iterations. These advantages combined provide a complete framework for sample-efficient off-policy optimization of movement primitives for robot learning of high-dimensional manipulation skills. Our experimental results conducted both in simulation and on a real robot show that LAMPO provides sample-efficient policies against common approaches in literature. Code available at https://github.com/SamuelePolimi/lampo.

show abstract

Prediction of Human Whole-Body Movements with AE- ProMPs

Cited by 6 publications

References 16 publications

Latent Ergonomics Maps: Real-Time Visualization of Estimated Ergonomics of Human Movements

Latent Ergonomics Maps: Real-Time Visualization of Estimated Ergonomics of Human Movements

Human-Humanoid Interaction and Cooperation: a Review

Contextual Latent-Movements Off-Policy Optimization for Robotic Manipulation Skills

Contact Info

Product

Resources

About