The Visual Neuroscience of Robotic Grasping

Chinellato, Eris; Pobil, Ángel P. del

doi:10.1007/978-3-319-20303-4

Cited by 23 publications

(4 citation statements)

References 203 publications

(254 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…I start by noting the many theoretical arguments offered in favor of dorsal/ventral interactions (Briscoe, 2009;Chinellato & Del Pobil, 2016;Ferretti, 2016Ferretti, , 2019McIntosh & Lashley, 2008;Zanon et al, 2010). These are often based on the idea that ventral-to-dorsal influences will be required for many actions, since just the ventral stream processes certain objectattributes required for the everyday interactions that we have with realworld objects (e.g., in the case of a coffee mug, things like "its weight, how full the cup is, and [its] temperature" (Briscoe, 2009, p. 431)).…”

Section: Interstream Interactionsmentioning

confidence: 99%

“…These are often based on the idea that ventral-to-dorsal influences will be required for many actions, since just the ventral stream processes certain objectattributes required for the everyday interactions that we have with realworld objects (e.g., in the case of a coffee mug, things like "its weight, how full the cup is, and [its] temperature" (Briscoe, 2009, p. 431)). In similar spirit, Chinellato and Del Pobil (2016) offer a detailed theoretical model of how visuomotor action unfolds in the primate visual system, one in which the ventral stream influences dorsal processing in a few different ways. 22 One of their more relevant hypotheses is the idea that the (ventrally located) lateral occipital cortex (LO) influences activity in the (dorsally located) anterior intraparietal sulcus (AIP), providing AIP with information about things such as an object's estimated size, as learned from previous experience with that object (2016, Chap.…”

Section: Interstream Interactionsmentioning

confidence: 99%

See 1 more Smart Citation

Conscious vision guides motor action—rarely

Kozuch

2022

Philosophical Psychology

View full text Add to dashboard Cite

According to Milner and Goodale's dual visual systems (DVS) theory, a division obtains between visual consciousness and motor action, in that the visual system producing conscious vision (the ventral stream) is distinct from the one guiding action (the dorsal stream). That there would be this division is often taken (by Andy Clark and others) to undermine the folk view on how consciousness and action relate. However, even if this division obtains, this leaves open the possibility that conscious ventral information is often transmitted to the unconscious dorsal stream and then used to guide action, a possibility seeming to preserve a significant role for consciousness in action. This article assesses this possibility. In course of doing so, we will review those arguments recently having been made against the DVS view on how visual consciousness and action relate (ones due to, e.g., Briscoe and Schwenkler, or Schenk and McIntosh). What we will find is that, if we properly analyze the data upon which these arguments are based, we are still left with the impression that the DVS view is largely correct; i.e., it is only rarely that visual experience guides action.

show abstract

Section: Interstream Interactionsmentioning

confidence: 99%

Section: Interstream Interactionsmentioning

confidence: 99%

Conscious vision guides motor action—rarely

Kozuch

2022

Philosophical Psychology

View full text Add to dashboard Cite

show abstract

“…Biological inspiration in robotics has been growing over the years with significant interaction between robotics and neuroscience, even though the neuroscience background is usually an endless inspiration with small impact on the final application [3]. Recent research on visuotactile grasping and manipulation has shown that robots have improved manipulation skills when simulating human tactile and visual perception to action.…”

Section: Literature Reviewmentioning

confidence: 99%

“…Grasping and manipulating objects is a distinctive part of the human-being skill set. It is an ability evolved from the erect posture that freed our upper limbs, turning our hands into two sophisticated sets of tools [3]. Not surprisingly, humans’ hand dexterity and reasoning are the holy grail of bio-inspired robotics control and actuation.…”

Section: Introductionmentioning

confidence: 99%

Estimating the Orientation of Objects from Tactile Sensing Data Using Machine Learning Methods and Visual Frames of Reference

Fonseca

Oliveira

Petriu

2019

Sensors

View full text Add to dashboard Cite

Underactuated hands are useful tools for robotic in-hand manipulation tasks due to their capability to seamlessly adapt to unknown objects. To enable robots using such hands to achieve and maintain stable grasping conditions even under external disturbances while keeping track of an in-hand object’s state requires learning object-tactile sensing data relationships. The human somatosensory system combines visual and tactile sensing information in their “What and Where” subsystem to achieve high levels of manipulation skills. The present paper proposes an approach for estimating the pose of in-hand objects combining tactile sensing data and visual frames of reference like the human “What and Where” subsystem. The system proposed here uses machine learning methods to estimate the orientation of in-hand objects from the data gathered by tactile sensors mounted on the phalanges of underactuated fingers. While tactile sensing provides local information about objects during in-hand manipulation, a vision system generates egocentric and allocentric frames of reference. A dual fuzzy logic controller was developed to achieve and sustain stable grasping conditions autonomously while forces were applied to in-hand objects to expose the system to different object configurations. Two sets of experiments were used to explore the system capabilities. On the first set, external forces changed the orientation of objects while the fuzzy controller kept objects in-hand for tactile and visual data collection for five machine learning estimators. Among these estimators, the ridge regressor achieved an average mean squared error of 0.077 ∘ . On the second set of experiments, one of the underactuated fingers performed open-loop object rotations and data recorded were supplied to the same set of estimators. In this scenario, the Multilayer perceptron (MLP) neural network achieved the lowest mean squared error of 0.067 ∘ .

show abstract