Open Access. Powered by Scholars. Published by Universities.®
Articles 1 - 8 of 8
Full-Text Articles in Engineering
Human Action Segmentation Based On A Streaming Uniform Entropy Slice Method, Cheng Peng, Sio-Long Lo, Jie Huang, Ah Chung Tsoi
Human Action Segmentation Based On A Streaming Uniform Entropy Slice Method, Cheng Peng, Sio-Long Lo, Jie Huang, Ah Chung Tsoi
Faculty of Engineering and Information Sciences - Papers: Part B
Segmentation of human actions is a major research problem in video understanding. A number of existing approaches demonstrate that performing action segmentation before action recognition results in better recognition performance. In this paper, we address the problem of action segmentation in an online manner. We first extend the clustering-based image segmentation approach into a temporal one, where hierarchical supervoxel levels for action segmentation are generated accordingly. We then propose a streaming approach to flatten the hierarchical levels into one based on Uniform Entropy Slice (UES), in order to preserve important information in the video. The flattened level contains the silhouette …
Motion Optimization And Parameter Identification For A Human And Lower Back Exoskeleton Model, Paul Manns, Manish Sreenivasa, Matthew Millard, Katja Mombaur
Motion Optimization And Parameter Identification For A Human And Lower Back Exoskeleton Model, Paul Manns, Manish Sreenivasa, Matthew Millard, Katja Mombaur
Faculty of Engineering and Information Sciences - Papers: Part B
Designing an exoskeleton to reduce the risk of low-back injury during lifting is challenging. Computational models of the human-robot system coupled with predictive movement simulations can help to simplify this design process. Here, we present a study that models the interaction between a human model actuated by muscles and a lower back exoskeleton. We provide a computational framework for identifying the spring parameters of the exoskeleton using an optimal control approach and forward-dynamics simulations. This is applied to generate dynamically consistent bending and lifting movements in the sagittal plane. Our computations are able to predict motions and forces of the …
Modeling And Identification Of A Realistic Spiking Neural Network And Musculoskeletal Model Of The Human Arm, And An Application To The Stretch Reflex, Manish Sreenivasa, Ko Ayusawa, Yoshihiko Nakamura
Modeling And Identification Of A Realistic Spiking Neural Network And Musculoskeletal Model Of The Human Arm, And An Application To The Stretch Reflex, Manish Sreenivasa, Ko Ayusawa, Yoshihiko Nakamura
Faculty of Engineering and Information Sciences - Papers: Part B
This study develops a multi-level neuromuscular model consisting of topological pools of spiking motor, sensory and interneurons controlling a bi-muscular model of the human arm. The spiking output of motor neuron pools were used to drive muscle actions and skeletal movement via neuromuscular junctions. Feedback information from muscle spindles were relayed via monosynaptic excitatory and disynaptic inhibitory connections, to simulate spinal afferent pathways. Subject-specific model parameters were identified from human experiments by using inverse dynamics computations and optimization methods. The identified neuromuscular model was used to simulate the biceps stretch reflex and the results were compared to an independent dataset. …
Walking Paths To And From A Goal Differ: On The Role Of Bearing Angle In The Formation Of Human Locomotion Paths, Manish Sreenivasa, Katja Mombaur, Jean-Paul Laumond
Walking Paths To And From A Goal Differ: On The Role Of Bearing Angle In The Formation Of Human Locomotion Paths, Manish Sreenivasa, Katja Mombaur, Jean-Paul Laumond
Faculty of Engineering and Information Sciences - Papers: Part B
The path that humans take while walking to a goal is the result of a cognitive process modulated by the perception of the environment and physiological constraints. The path shape and timing implicitly embeds aspects of the architecture behind this process. Here, locomotion paths were investigated during a simple task of walking to and from a goal, by looking at the evolution of the position of the human on a horizontal (x,y) plane. We found that the path while walking to a goal was not the same as that while returning from it. Forward-return paths were systematically separated by 0.5-1.9m, …
Hrp-2 Plays The Yoyo: From Human To Humanoid Yoyo Playing Using Optimal Control, Katja Mombaur, Manish Sreenivasa
Hrp-2 Plays The Yoyo: From Human To Humanoid Yoyo Playing Using Optimal Control, Katja Mombaur, Manish Sreenivasa
Faculty of Engineering and Information Sciences - Papers: Part B
Yoyo playing may seem easy for a human, but it is a challenging problem for a humanoid robot. This paper presents an approach to generate yoyo motions for the humanoid robot, HRP-2, based on motion recorded from human yoyo playing, dynamical models and numerical optimal control techniques. We recorded vertical yoyo playing of 4 subjects measuring yoyo height and rotation angle as well as the corresponding hand motions. A detailed multi-phase yoyo model with impact collisions and control patterns of human yoyo playing were identified from these measurements. Based on this knowledge, reliable yoyo motions within the feasibility ranges of …
Inverse Optimal Control As A Tool To Understand Human Yoyo Playing, Katja Mombaur, Manish Sreenivasa
Inverse Optimal Control As A Tool To Understand Human Yoyo Playing, Katja Mombaur, Manish Sreenivasa
Faculty of Engineering and Information Sciences - Papers: Part B
This paper presents an inverse optimal control approach to identify objective functions of human motion from motion capture measurements. We apply it to analyze human yoyo playing. Yoyo playing may seem easy to us to learn but it is a challenging problem from a mechanical point of view involving a hybrid dynamics model. We recorded vertical yoyo playing of humans measuring yoyo height and rotation angle as well as the corresponding hand motions. Results of inverse optimal control are presented showing a mixed criterion of cycle time and terms depending on yoyo and hand acceleration and velocity.
On Using Human Movement Invariants To Generate Target-Driven Anthropomorphic Locomotion, Manish Sreenivasa, Philippe Souères, Jean-Paul Laumond
On Using Human Movement Invariants To Generate Target-Driven Anthropomorphic Locomotion, Manish Sreenivasa, Philippe Souères, Jean-Paul Laumond
Faculty of Engineering and Information Sciences - Papers: Part B
We present a method for generating anthropomorphic motion by studying `invariants' in human movements and applying them as kinematic tasks. We recorded whole-body motion of 14 participants during a walking and grasping task and performed a detailed analysis in order to synthesize the stereotypy in human motion as rules. We propose an algorithm that produces the key parameters of motion taking into account the knowledge from human movement, and the limitations of the anthropomorph. We generalize our results such that we can create motion parameters for objects which were not in the original protocol. The algorithmic output is applied in …
On Real-Time Whole-Body Human To Humanoid Motion Transfer, Francisco-Javier Montecillo-Puente, Manish Sreenivasa, Jean-Paul Laumond
On Real-Time Whole-Body Human To Humanoid Motion Transfer, Francisco-Javier Montecillo-Puente, Manish Sreenivasa, Jean-Paul Laumond
Faculty of Engineering and Information Sciences - Papers: Part B
We present a framework for online imitation of human motion by the humanoid robot HRP-2. We introduce a representation of human motion, the humanoid-Normalized model, and a Center of Mass (CoM) anticipation model to prepare the robot in case the human lifts his/her foot. Our proposed motion representation codifies operational space and geometric information. Whole body robot motion is computed using a task-based prioritized inverse kinematics solver. By setting the human motion model as the target, and giving the maintenance of robot CoM a high priority, we can achieve a large range of motion imitation. We present two scenarios of …