DATA REPORT article Front. Neurorobot., 18 June 2019 Volume 13 - | https://doi.org/10.3389/fnbot.2019.00038

The hand gesture recognition system is a noncontact and intuitive communication approach, which, in turn, allows for natural efficient interaction. This work focuses on developing novel robust system, which insensitive to environmental illumination background variation. In the field of recognition, standard vision sensors, such as CMOS cameras, are widely used sensing devices state-of-the-art systems. However, cameras depend constraints, lighting variability cluttered background,...

Robot manipulation in cluttered environments of-ten requires complex and sequential rearrangement of multiple objects order to achieve the desired reconfiguration target objects. Due sophisticated physical interactions involved such scenarios, rearrangement-based is still limited a small range tasks especially vulnerable uncertainties perception noise. This paper presents planning framework that leverages efficiency sampling-based approaches, closes loop by dynamically controlling horizon....

In order to successfully perform manipulation tasks in new environments, such as grasping, robots must be proficient segmenting unseen objects from the background and/or other objects. Previous works object instance segmentation (UOIS) by training deep neural networks on large-scale data learn RGB/RGB-D feature embeddings, where cluttered environments often result inaccurate segmentations. We build upon these methods and introduce a novel approach correct segmentation, under-segmentation, of...

Nonprehensile manipulation through precise pushing is an essential skill that has been commonly challenged by perception and physical uncertainties, such as those associated with contacts, object geometries, properties. For this, we propose a unified framework jointly addresses system modeling, action generation, control. While most existing approaches either heavily rely on priori information for analytic or leverage large dataset to learn dynamic models, our approximates transition...

Calibrating robots into their workspaces is crucial for manipulation tasks. Existing calibration techniques often rely on sensors external to the robot (cameras, laser scanners, etc.) or specialized tools. This reliance complicates process and increases costs time requirements. Furthermore, associated setup measurement procedures require significant human intervention, which makes them more challenging operate. Using built-in force-torque sensors, are nowadays a default component in...

Nonprehensile actions such as pushing are crucial for addressing multi-object rearrangement problems. To date, existing nonprehensile solutions all robot-centric, i.e., the manipulation generated with robot-relevant intent and their outcomes passively evaluated afterwards. Such pipelines very different from human strategies typically inefficient. this end, work proposes a novel object-centric planning paradigm develops first planner general By assuming that each object can actively move...

Real-world object manipulation has been commonly challenged by physical uncertainties and perception limitations. Being an effective strategy, while caging configuration-based frameworks have successfully provided robust solutions, they are not broadly applicable due to their strict requirements on the availability of multiple robots, widely distributed contacts, or specific geometries robots objects. To this end, work proposes a novel concept, termed Caging in Time, allow configurations be...

Traditional robotic manipulation mostly focuses on collision-free tasks. In practice, however, many tasks (e.g., occluded object grasping) require the robot to intentionally collide with environment reach a desired task configuration. By enabling compliant motions, collisions between and are allowed can thus be exploited, but more physical uncertainties introduced. To address collision-rich problems such as grasping while handling involved uncertainties, we propose collision-inclusive...

Rearrangement-based nonprehensile manipulation still remains as a challenging problem due to the high-dimensional space and complex physical uncertainties it entails. We formulate this class of problems coupled local rearrangement global action optimization by incorporating free-space transit motions between constrained rearranging actions. propose forest-based kinodynamic planning framework concurrently search in multiple regions, so enable exploration most task-relevant subspaces, while...

Building hand-object models for dexterous in-hand manipulation remains a crucial and open problem. Major challenges include the difficulty of obtaining geometric dynamical hand, object, time-varying contacts, as well inevitable physical perception uncertainties. Instead building accurate to map between actuation inputs object motions, this work proposes enable systems continuously approximate their local via self-identification process where an underlying model is estimated through small...

Integrating Internet of things (IoT) techniques into automated vehicles has been a vision in intelligent transportation system, there is however seldom researches addressing it. To this end, we envision scenario: short‐range on‐board sensor perception system attached to individual mobile applications such as are connected via IoT and transferred long‐range mobile‐sensing which can be used part more extensive surveilling the environment. However, sensing brings new challenges for how...

This paper investigates trajectory generation alternatives for creating single-stroke light paintings with a small quadrotor robot. We propose to reduce the cost of minimum snap piecewise polynomial passing through set waypoints by displacing those towards or away from camera while preserving their projected position. It is in regions high curvature, where are close together, that we make modifications snap, and evaluate two different strategies: one uses full range depths increase distance...

Building hand-object models for dexterous in-hand manipulation remains a crucial and open problem. Major challenges include the difficulty of obtaining geometric dynamical hand, object, time-varying contacts, as well inevitable physical perception uncertainties. Instead building accurate to map between actuation inputs object motions, this work proposes enable systems continuously approximate their local via self-identification process where an underlying model is estimated through small...

Rearrangement-based nonprehensile manipulation still remains as a challenging problem due to the high-dimensional space and complex physical uncertainties it entails. We formulate this class of problems coupled local rearrangement global action optimization by incorporating free-space transit motions between constrained rearranging actions. propose forest-based kinodynamic planning framework concurrently search in multiple regions, so enable exploration most task-relevant subspaces, while...

Robot manipulation in cluttered environments often requires complex and sequential rearrangement of multiple objects order to achieve the desired reconfiguration target objects. Due sophisticated physical interactions involved such scenarios, rearrangement-based is still limited a small range tasks especially vulnerable uncertainties perception noise. This paper presents planning framework that leverages efficiency sampling-based approaches, closes loop by dynamically controlling horizon....