scholarly journals Sensorimotor adaptation compensates for distortions of 3D shape information

2019 ◽  
Author(s):  
Evan Cesanek ◽  
Jordan A. Taylor ◽  
Fulvio Domini
Keyword(s):  
Visual Processing ◽  
Sensory Feedback ◽  
Action Planning ◽  
Physical Reality ◽  
Depth Information ◽  
Sensorimotor Adaptation ◽  
Shape Information ◽  
3D Shape ◽  
Depth Cue ◽  
3D Information

AbstractVisual perception often fails to recover the veridical 3D shape of objects in the environment due to ambiguity and variability in the available depth cues. However, we rely heavily on 3D shape estimates when planning movements, for example reaching to pick up an object from a slanted surface. Given the wide variety of distortions that can affect 3D perception, how do our actions remain accurate across different environments? One hypothesis is that the visuomotor system performs selective filtering of 3D information to minimize distortions. Indeed, some studies have found that actions appear to preferentially target stereo information when it is put in conflict with texture information. However, since these studies analyze averages over multiple trials, this apparent preference could be produced by sensorimotor adaptation. In Experiment 1, we create a set of cue-conflict stimuli where one available depth cue is affected by a constant bias. Sensory feedback rapidly aligns the motor output with physical reality in just a few trials, which can make it seem as if action planning selectively relies on the reinforced cue—yet no change in the relative influences of the cues is necessary to eliminate the constant errors. In contrast, when one depth cue becomes less correlated with physical reality, variable movement errors will occur, causing canonical adaptation to fail as the opposite error corrections cancel out. As a result, canonical adaptation cannot explain the preference for stereo found in studies with variable errors. However, Experiment 2 shows that persistent errors can produce a novel form of adaptation that gradually reduces the relative influence of an unreliable depth cue. These findings show that grasp control processes are continuously modified based on sensory feedback to compensate for both biases and noise in 3D visual processing, rather than having a hardwired preference for one type of depth information.

scholarly journals Sensorimotor adaptation and cue reweighting compensate for distorted 3D shape information, accounting for paradoxical perception-action dissociations

2020 ◽  
Vol 123 (4) ◽  
pp. 1407-1419
Author(s):  
Evan Cesanek ◽  
Jordan A. Taylor ◽  
Fulvio Domini
Keyword(s):  
Supervised Learning ◽  
Sensory Feedback ◽  
Perception And Action ◽  
Three Dimensional ◽  
Learning Processes ◽  
Sensorimotor Adaptation ◽  
Slant Perception ◽  
Fixed Amount ◽  
Shape Information ◽  
3D Shape

Visually guided movements can show surprising accuracy even when the perceived three-dimensional (3D) shape of the target is distorted. One explanation of this paradox is that an evolutionarily specialized “vision-for-action” system provides accurate shape estimates by relying selectively on stereo information and ignoring less reliable sources of shape information like texture and shading. However, the key support for this hypothesis has come from studies that analyze average behavior across many visuomotor interactions where available sensory feedback reinforces stereo information. The present study, which carefully accounts for the effects of feedback, shows that visuomotor interactions with slanted surfaces are actually planned using the same cue-combination function as slant perception and that apparent dissociations can arise due to two distinct supervised learning processes: sensorimotor adaptation and cue reweighting. In two experiments, we show that when a distorted slant cue biases perception (e.g., surfaces appear flattened by a fixed amount), sensorimotor adaptation rapidly adjusts the planned grip orientation to compensate for this constant error. However, when the distorted slant cue is unreliable, leading to variable errors across a set of objects (i.e., some slants are overestimated, others underestimated), then relative cue weights are gradually adjusted to reduce the misleading effect of the unreliable cue, consistent with previous perceptual studies of cue reweighting. The speed and flexibility of these two forms of learning provide an alternative explanation of why perception and action are sometimes found to be dissociated in experiments where some 3D shape cues are consistent with sensory feedback while others are faulty. NEW & NOTEWORTHY When interacting with three-dimensional (3D) objects, sensory feedback is available that could improve future performance via supervised learning. Here we confirm that natural visuomotor interactions lead to sensorimotor adaptation and cue reweighting, two distinct learning processes uniquely suited to resolve errors caused by biased and noisy 3D shape cues. These findings explain why perception and action are often found to be dissociated in experiments where some cues are consistent with sensory feedback while others are faulty.

scholarly journals Explicit and implicit depth-cue integration: evidence of systematic biases with real objects

2021 ◽  
Author(s):  
Carlo Campagnoli ◽  
Bethany Hung ◽  
Fulvio Domini
Keyword(s):  
3D Structure ◽  
Normative Theory ◽  
Matching Task ◽  
Depth Information ◽  
Cue Integration ◽  
Shape Estimation ◽  
Depth Cues ◽  
3D Shape ◽  
Real Objects ◽  
Depth Cue

AbstractIn a previous series of experiments using virtual stimuli, we found evidence that 3D shape estimation agrees to a superadditivity rule of depth-cue combination. According to this rule, adding depth cues leads to greater perceived depth magnitudes and, in principle, to depth overestimation. The mechanism underlying the superadditivity effect can be fully accounted for by a normative theory of cue integration, through the adaptation of a model of cue integration termed the Intrinsic Constraint (IC) model. As for its nature, it remains unclear whether superadditivity is a byproduct of the artificial nature of virtual environments, causing explicit reasoning to infiltrate behavior and inflate the depth judgments when a scene is richer in depth cues, or the genuine output of the process of depth-cue integration. In the present study, we addressed this question by testing whether the IC model’s prediction of superadditivity generalizes beyond VR environments to real world situations. We asked participants to judge the perceived 3D shape of cardboard prisms through a matching task. To assay the potential influence of explicit control over those perceptual estimates, we also asked participants to reach and hold the same objects with their fingertips and we analyzed the in-flight grip size during the reaching. Using physical objects ensured that all visual information was fully consistent with the stimuli’s 3D structure without computer-generated artifacts. We designed a novel technique to carefully control binocular and monocular 3D cues independently from one another, allowing to add or remove depth information from the scene seamlessly. Even with real objects, participants exhibited a clear superadditivity effect in both explicit and implicit tasks. Furthermore, the magnitude of this effect was accurately predicted by the IC model. These results confirm that superadditivity is an inherent feature of depth estimation.

scholarly journals Neural coding of action planning: visual processing or visual memory?

2016 ◽  
Vol 16 (12) ◽  
pp. 23
Author(s):  
Simona Monaco ◽  
Elisa Pellencin ◽  
Malfatti Giulia ◽  
Turella Luca
Keyword(s):  
Visual Processing ◽  
Neural Coding ◽  
Visual Memory ◽  
Action Planning

Contour Interaction in Visual Space

1978 ◽  
Vol 22 (1) ◽  
pp. 74-77
Author(s):  
Robert Fox
Keyword(s):  
Visual Processing ◽  
Visual Masking ◽  
Random Element ◽  
Visual Space ◽  
Depth Information ◽  
Implicit Assumption ◽  
Depth Position ◽  
Series Of Experiments ◽  
Contour Interaction ◽  
Inhibitory Interactions

Virtually all the extensive research on inhibitory interactions among adjacent visual stimuli seen in such phenomena as simultaneous contrast and visual masking have employed situations in which the interacting stimulus elements occupy the same depth plane, i.e., the z-axis values are the same, in deference to the implicit assumption that processing of depth information occurs only after the visual processing of contour information is completed. But there are theoretical reasons and some data suggesting that the interactions among contours depend critically upon their relative positions in depth—interactions may not occur if the stimulus elements occupy different depth positions. The extent to which the metacontrast form of visual masking is dependent upon depth position was investigated in a series of experiments that used stereoscopic contours formed from random-element stereograms as test and mask stimuli. The random-element stereogram generation system permitted large variations in depth to be made without introducing confounding changes in proximal stimulation. The main results are 1) separation of test and mask stimuli in depth substantially reduces masking, and 2) when more than one stimulus is in visual space the stimulus that either appears first or appears closer to the observer receives preferential processing by the visual system.

scholarly journals Analysis of RGB-D camera technologies for supporting different facial usage scenarios

2020 ◽  
Vol 79 (39-40) ◽  
pp. 29375-29398
Author(s):  
Luca Ulrich ◽  
Enrico Vezzetti ◽  
Sandro Moos ◽  
Federica Marcolin
Keyword(s):  
Correct Choice ◽  
Depth Information ◽  
Expression Recognition ◽  
Face Expression ◽  
Starting Point ◽  
3D Camera ◽  
Face Expression Recognition ◽  
3D Information ◽  
Face Morphology ◽  
New Algorithms

Abstract Recently a wide variety of applications has been developed integrating 3D functionalities. Advantages given by the possibility of relying on depth information allows the developers to design new algorithms and to improve the existing ones. In particular, for what concerns face morphology, 3D has led to the possibility to obtain face depth maps highly close to reality and consequently an improvement of the starting point for further analysis such as Face Detection, Face Authentication, Face Identification and Face Expression Recognition. The development of the aforementioned applications would have been impossible without the progress of sensor technologies for obtaining 3D information. Several solutions have been adopted over time. In this paper, emphasis is put on passive stereoscopy, structured light, time-of-flight (ToF) and active stereoscopy, namely the most used technologies for the cameras design and fulfilment according to the literature. The aim of this article is to investigate facial applications and to examine 3D camera technologies to suggest some guidelines for addressing the correct choice of a 3D sensor according to the application that has to be developed.

On the Stereoscopic Composition of Wide Baseline Stereo Pairs

2009 ◽  
Author(s):  
Chensheng Wang ◽  
Xiaochun Wang ◽  
Joris S. M. Vergeest ◽  
Tjamme Wiegers
Keyword(s):  
Binocular Vision ◽  
Main Idea ◽  
Depth Information ◽  
Stereoscopic Image ◽  
Stereoscopic Images ◽  
Wide Baseline Stereo ◽  
Before And After ◽  
Depth Cue ◽  
A Minor

Wide baseline cameras are broadly utilized in binocular vision systems, delivering depth information and stereoscopic images of the scene that are crucial both in virtual reality and in computer vision applications. However, due to the large distance between the two cameras, the stereoscopic composition of stereo pairs with wide baseline is hardly to fit the human eye parallax. In this paper, techniques and algorithms for the stereoscopic composition of wide baseline stereo pairs in binocular vision will be investigated. By incorporating the human parallax limitation, a novel algorithm being capable of adjusting the wide baseline stereo pairs to compose a high quality stereoscopic image will be formulated. The main idea behind the proposed algorithm is, by simulating the eyeball rotation, to shift the wide baseline stereo pairs closer to each other to fit the human parallax limit. This makes it possible for the wide baseline stereo pairs to be composed into a recognizable stereoscopic image in terms of human parallax with a minor cost of variation in the depth cue. In addition, the depth variations before and after the shifting of the stereo pairs are evaluated by conducting an error estimation. Examples are provided for the evaluation of the proposed algorithm. And the quality of the composed stereoscopic images proves that the proposed algorithm is both valid and effective.

scholarly journals Hybrid 3D Shape Measurement Using the MEMS Scanning Micromirror

Micromachines ◽  
2019 ◽  
Vol 10 (1) ◽  
pp. 47 ◽  
Author(s):  
Tao Yang ◽  
Guanliang Zhang ◽  
Huanhuan Li ◽  
Xiang Zhou
Keyword(s):  
Laser Beam ◽  
Microelectromechanical Systems ◽  
Laser Source ◽  
Shape Measurement ◽  
Fusion Algorithm ◽  
Projection System ◽  
3D Shape ◽  
3D Shape Measurement ◽  
Complete Measurement ◽  
3D Information

A surface with large reflection variations represents one of the biggest challenges for optical 3D shape measurement. In this work, we propose an alternative hybrid 3D shape measurement approach, which combines the high accuracy of fringe projection profilometry (FPP) with the robustness of laser stripe scanning (LSS). To integrate these two technologies into one system, first, we developed a biaxial Microelectromechanical Systems (MEMS) scanning micromirror projection system. In this system, a shaped laser beam serves as a light source. The MEMS micromirror projects the laser beam onto the object surface. Different patterns are produced by controlling the laser source and micromirror jointly. Second, a quality wised algorithm is delivered to develop a hybrid measurement scheme. FPP is applied to obtain the main 3D information. Then, LSS helps to reconstruct the missing depth guided by the quality map. After this, the data fusion algorithm is used to merge and output complete measurement results. Finally, our experiments show significant improvement in the accuracy and robustness of measuring a surface with large reflection variations. In the experimental instance, the accuracy of the proposed method is improved by 0.0278 mm and the integrity is improved by 83.55%.

The Perception of Distance in Simulated Visual Displays:A Comparison of the Effectiveness and Accuracy of Multiple Depth Cues Across Viewing Distances

1997 ◽  
Vol 6 (5) ◽  
pp. 513-531 ◽  
Author(s):  
R. Troy Surdick ◽  
Elizabeth T. Davis ◽  
Robert A. King ◽  
Larry F. Hodges
Keyword(s):  
Relative Size ◽  
Linear Perspective ◽  
Depth Information ◽  
Texture Gradient ◽  
Viewing Distance ◽  
Relative Height ◽  
Distance Information ◽  
Depth Cues ◽  
Relative Brightness ◽  
Depth Cue

The ability effectively and accurately to simulate distance in virtual and augmented reality systems is a challenge currently facing R&D. To examine this issue, we separately tested each of seven visual depth cues (relative brightness, relative size, relative height, linear perspective, foreshortening, texture gradient, and stereopsis) as well as the condition in which all seven of these cues were present and simultaneously providing distance information in a simulated display. The viewing distances were 1 and 2 m. In developing simulated displays to convey distance and depth there are three questions that arise. First, which cues provide effective depth information (so that only a small change in the depth cue results in a perceived change in depth)? Second, which cues provide accurate depth information (so that the perceived distance of two equidistant objects perceptually matches)? Finally, how does the effectiveness and accuracy of these depth cues change as a function of the viewing distance? Ten college-aged subjects were tested with each depth-cue condition at both viewing distances. They were tested using a method of constant stimuli procedure and a modified Wheat-stone stereoscopic display. The perspective cues (linear perspective, foreshortening, and texture gradient) were found to be more effective than other depth cues, while effectiveness of relative brightness was vastly inferior. Moreover, relative brightness, relative height, and relative size all significantly decreased in effectiveness with an increase in viewing distance. The depth cues did not differ in terms of accuracy at either viewing distance. Finally, some subjects experienced difficulty in rapidly perceiving distance information provided by stereopsis, but no subjects had difficulty in effectively and accurately perceiving distance with the perspective information used in our experiment. A second experiment demonstrated that a previously stereo-anomalous subject could be trained to perceive stereoscopic depth in a binocular display. We conclude that the use of perspective cues in simulated displays may be more important than the other depth cues tested because these cues are the most effective and accurate cues at both viewing distances, can be easily perceived by all subjects, and can be readily incorporated into simpler, less complex displays (e.g., biocular HMDs) or more complex ones (e.g., binocular or see-through HMDs).

Sign in / Sign up

Export Citation Format

Share Document