Visual Anchor Research Articles

Enhancing Sound Source Localization via False Negative Elimination.

Sound source localization aims to localize objects emitting the sound in visual scenes. Recent works obtaining impressive results typically rely on contrastive learning. However, the common practice of randomly sampling negatives in prior arts can lead to the false negative issue, where the sounds semantically similar to visual instance are sampled as negatives and incorrectly pushed away from the visual anchor/query. As a result, this misalignment of audio and visual features could yield inferior performance. To address this issue, we propose a novel audio-visual learning framework which is instantiated with two individual learning schemes: self-supervised predictive learning (SSPL) and semantic-aware contrastive learning (SACL). SSPL explores image-audio positive pairs alone to discover semantically coherent similarities between audio and visual features, while a predictive coding module for feature alignment is introduced to facilitate the positive-only learning. In this regard SSPL acts as a negative-free method to eliminate false negatives. By contrast, SACL is designed to compact visual features and remove false negatives, providing reliable visual anchor and audio negatives for contrast. Different from SSPL, SACL releases the potential of audio-visual contrastive learning, offering an effective alternative to achieve the same goal. Comprehensive experiments demonstrate the superiority of our approach over the state-of-the-arts. Furthermore, we highlight the versatility of the learned representation by extending the approach to audio-visual event classification and object detection tasks. Code and models are available at: https://github.com/zjsong/SACL.

Relative foveal dark adaptation: a potential method for assessing macular health.

Dark adaptation measures photoreceptor recovery following intense light stimulation. Time to recovery reflects retinal function. We describe a novel method of relative foveal dark adaptation using an iPhone. Data from a small number of healthy subjects were studied to assess reproducibility, effects of age, and consider potential clinical utility. Relative foveal dark adaption was studied in 6 normal subjects across ages from 20 to 81 years and across differing testing conditions. Foveal bleaching is produced by fixating a bright white circle on an iPhone for variable times. After foveal bleaching an annular surround appears to complete a bullseye stimulus with surround initially brighter than centre. As the fovea recovers the centre regains brightness. Relative foveal dark adaptation, the time for the visual anchor to shift from surround to centre, was studied across a range of bleaching times, ages, and testing conditions. Dispersion of dark adaptation times grows with increasing age. Foveal bleaching for 30 s was as effective as longer times. Testing times with a 30 s bleach were less than 1 min. Foveal dark adaptation was reproducible within each subject and was unaffected by ambient room lighting, pupil size, and light attenuation. Repeat, immediately sequential testing was similarly reproducible except after long bleaching. This method of dark adaptation is intuitive, repeatable, and relatively unaffected by testing condition. Testing times are brief, requiring only an iPhone screen positioned at reading distance. Relative foveal dark adaptation may be a useful tool to assess macular health.

Microsaccade generation requires a foveal anchor.

Visual scene characteristics can affect various aspects of saccade and microsaccade dynamics. For example, blank visual scenes are known to elicit diminished saccade and microsaccade production, compared to natural scenes. Similarly, microsaccades are less frequent in the dark. Yet, the extent to which foveal versus peripheral visual information contribute to microsaccade production remains unclear: because microsaccade directions are biased towards covert attention locations, it follows that peripheral visual stimulation could suffice to produce regular microsaccade dynamics, even without foveal stimulation being present. Here we determined the characteristics of microsaccades as a function of foveal and/or peripheral visual stimulation, while human subjects conducted four types of oculomotor tasks (fixation, free viewing, guided viewing and passive viewing). Foveal information was either available, or made unavailable, by the presentation of simulated scotomas. We found foveal stimulation to be critical for microsaccade production, and peripheral stimulation, by itself, to be insufficient to yield normal microsaccades. In each oculomotor task, microsaccade production decreased when scotomas blocked foveal stimulation. Across comparable foveal stimulation conditions, the type of peripheral stimulation (static versus dynamic) moreover affected microsaccade production, with dynamic backgrounds resulting in lower microsaccadic rates than static backgrounds. These results indicate that a foveal visual anchor is necessary for normal microsaccade generation. Whereas peripheral visual stimulation, on its own, does not suffice for normal microsaccade production, it can nevertheless modulate microsaccadic characteristics. These findings extend our current understanding of the links between visual input and ocular motor control, and may therefore help improve the diagnosis and treatment of ophthalmic conditions that degrade central vision, such as age-related macular degeneration.

Supporting Mathematic Achievement for Students with Learning Disabilities Through Enhanced Anchored Instruction

Difficulties in mathematical learning are common and significant.The struggle has increased exponentially with the Common Core State Standards in effect. Students with or without learning or mathematical disabilities may respond to continued failure by withdrawing their effort, carrying low self-esteem, and/or displaying avoidance behaviors. Enhanced anchored instruction is designed to provide students the opportunity to construct knowledge and design solutions to problems collaboratively. The approach of enhanced anchored instruction attempts to guide students to become more active in learning through the use of technology. Enhanced anchored instruction provides students the opportunity to work through problems attached to a visual anchor. These anchors help develop higher-order thinking and contextualized learning. This article discusses the steps to use enhanced anchored instruction to assist students with learning disabilities to improve problem-solving skills and increase mathematical achievement.

The Garden Palace: Building an Early Sydney Icon

The Garden Palace: Building an Early Sydney Icon

Visual processing is biased in peripersonal foot space

Objects in peripersonal space are of great importance for interaction with the sensory world. A variety of research exploring sensory processing in peripersonal space has produced extensive evidence for altered vision near the hands. However, visual representations of the peripersonal space surrounding the feet remain unexplored. In a set of four experiments, we investigated whether observers experience biases in visual processing for objects near the feet that mirror the alterations associated with near-hand space. Participants performed attentional-cueing tasks in which they detected targets appearing (1) near or far from a single visible foot, (2) near one of two visible feet, (3) near or far from a nonfoot visual anchor, or (4) near or far from an occluded foot. We found a temporal cost associated with detecting targets appearing far from a visible foot, but no biases were associated with targets appearing near versus far from either a nonfoot visual anchor or an occluded foot. These results provide the first evidence suggesting that objects within stepping or kicking distance are processed differently from objects outside of peripersonal foot space.

Anchoring in experimental asset markets

We investigate the relationship between anchoring and the emergence of bubbles in experimental asset markets. We show that setting a visual anchor at the fundamental value (FV) in the first period only is sufficient to eliminate or to significantly reduce bubbles in laboratory asset markets. If no FV-anchor is set, bubble-crash patterns emerge. Our results indicate that bubbles in laboratory environments are primarily sparked in the first period. If prices are initiated around the FV, they stay close to the FV over the entire trading horizon. Our insights can be related to initial public offerings and the interaction between prices set on pre-opening markets and subsequent intra-day price dynamics.

When men appear smaller or larger than they really are: preliminary evidence that women are fooled by size illusions in attractiveness judgment tasks

In humans, studies have shown that contrast illusions can affect perceptions of facial attractiveness and dominance. In non-human animals, recent research found that contrast illusions of size positively affected male mate value. In humans, male height is a potentially important indicator of mate value, with women preferring men taller than themselves. We tested in two studies whether height contrast illusions could affect women’s perceptions of male height and mate value, particularly attractiveness, dominance, and muscularity. Using computer-generated images of men of different heights standing in groups of three, 104 female participants rated targets either surrounded by shorter, same height, or taller distractors in a within-subject design. The second experiment (N=80) replicated and extended the first by making the images more realistic and adding natural backgrounds, suggesting that when participants are given a visual anchor, in order to get a better sense of the absolute height of the targets, the effects remain. In both studies, results showed that, compared with same height distractors, male targets were rated as taller when surrounded by shorter distractors, and as shorter when surrounded by taller distractors. Additionally, attractiveness, dominance, and muscularity perceptions were affected in a similar manner, with most of the differences in these appraisals being mediated by the perceived height differences. Therefore, differently sized distractors affected the perceived height and mate value of the targets, which were in effect all of the same constant size. These findings indicate that context dependent effects could potentially influence attractiveness judgments. The same man might thus be perceived as more attractive when surrounded by men of similar or smaller height, as opposed to when surrounded by men who are taller.

Anchoring in Experimental Asset Markets

We investigate the relationship between anchoring and the emergence of bubbles in experimental asset markets. We show that setting a visual anchor at the fundamental value (FV) in the first period only is sufficient to eliminate or to significantly reduce bubbles in laboratory asset markets. If no FV-anchor is set, bubble-crash patterns emerge. Our results indicate that bubbles in laboratory environments are primarily sparked in the first period. If prices are initiated around the FV, they stay close to the FV over the entire trading horizon. Our insights can be related to initial public offerings and the interaction between prices set on pre-opening markets and subsequent intra-day price dynamics.

Does calculating impair postural stabilization allowed by visual cues?

In many daily situations, balance control is associated with a cognitive activity such as reading or a simple calculation. The objective of this study was to investigate the relationship between these two specific human activities, especially the influence of visual cues and support surface stability on body sway during a calculation task. A Sensory Organization Test, which can disrupt or suppress sensory inputs, was performed on 71 healthy young adults. The evaluations were performed both with and without mental arithmetic tasks which consisted of backward counting by three or thirteen. Our results showed that the addition of a calculation task induced an increase in body sway only when visual cues were available. They also showed the same instability effect of the support surface on the amount of body sway no matter what the associated cognitive task was. Moreover, no difference in body sway was observed between the two calculation tasks no matter what the visual context and/or the stability of the support surface were. We suggest that focusing on fulfilling the requirements of the mental calculation challenge may be responsible for the increase in body sway. This increase may be related to the use of oculomotor activity as unintentional attempts to increase arousal by self-generated body movement. Thus, this activity facilitates information processing rather than minimizing unbalance by a visual anchor point.

Flight Test Results of a 2D Snapshot Hover

The viability of using visual snapshots to sense and control the hover of a Micro Air Vehicle (MAV) is considered. A new algorithm is proposed that uses a stored image or 'snapshot' of the ground directly under the MAV as a visual anchor point. For hover control, adjacent frames are used to calculate the velocity of the MAV for providing speed feedback and the absolute position error feedback of the MAV is calculated by directly comparing the subsequent frames with the stored snapshot. Flight tests in both indoor and outdoor environments demonstrate the ability of the proposed method to effectively check long-term drift of the vehicle on well textured ground against light wind disturbance.

Perifoveal spatial compression

We report a strong compression of space around a visual anchor presented in the near visual periphery (<5°). While subjects kept fixation, a salient visual stimulus (from now on referred to as "anchor") was presented, followed by a brief whole-field mask. At various times around mask onset a probe dot was flashed. Subjects estimated the position of the probe dot in relation to a subsequently presented comparison bar. The probe dot location was perceived nearly veridically when presented long before or after mask onset. However, when the probe dot was presented simultaneously with the mask it appeared shifted toward the anchor by as much as 50% of their separation. The anchor had to appear briefly before mask onset to attract the probe dot. No compression occurred when the anchor was presented long before or after the mask. When the probe dot and anchor were presented with similar brief duration, the more peripheral stimulus always shifted toward the more foveal stimulus independently of their temporal order. We hypothesize that the attraction might be explained by the summation of the neural activity distributions of probe and anchor.

Balancing on a Slackline: 8-Year-Olds vs. Adults

Children are less stable than adults during static upright stance. We investigated whether the same holds true for a task that was novel for both children and adults and highly dynamic: single-legged stance on a slackline. We compared 8-year-olds with young adults and assessed the following outcome measures: time on the slackline, stability on the slackline (calculated from slackline reaction force), gaze movement, head-in-space rotation and translation, trunk-in-space rotation, and head-on-trunk rotation. Eight-year-olds fell off the slackline quicker and were generally less stable on the slackline than adults. Eight-year-olds also showed more head-in-space rotation and translation, and more gaze variability around a visual anchor point they were instructed to fixate. Trunk-in-space and head-on-trunk rotations did not differ between groups. The results imply that the lower postural stability of 8-year-olds compared to adults – as found in simple upright stance – holds true for dynamic, novel tasks in which adults lack the advantage of more practice. They also suggest that the lack of head and gaze stability constitutes an important limiting factor in children’s ability to master such tasks.

Portion size me: Plate-size induced consumption norms and win-win solutions for reducing food intake and waste.

Research on the self-serving of food has empirically ignored the role that visual consumption norms play in determining how much food we serve on different sized dinnerware. We contend that dinnerware provides a visual anchor of an appropriate fill-level, which in turn, serves as a consumption norm (Study 1). The trouble with these dinnerware-suggested consumption norms is that they vary directly with dinnerware size--Study 2 shows Chinese buffet diners with large plates served 52% more, ate 45% more, and wasted 135% more food than those with smaller plates. Moreover, education does not appear effective in reducing such biases. Even a 60-min, interactive, multimedia warning on the dangers of using large plates had seemingly no impact on 209 health conference attendees, who subsequently served nearly twice as much food when given a large buffet plate 2 hr later (Study 3). These findings suggest that people may have a visual plate-fill level--perhaps 70% full--that they anchor on when determining the appropriate consumption norm and serving themselves. Study 4 suggests that the Delboeuf illusion offers an explanation why people do not fully adjust away from this fill-level anchor and continue to be biased across a large range of dishware sizes. These findings have surprisingly wide-ranging win-win implications for the welfare of consumers as well as for food service managers, restaurateurs, packaged goods managers, and public policy officials.

Design of a 3D snapshot based visual flight control system using a single camera in hover

The problem of developing a reliable system for sensing and controlling the hover of a Micro Air Vehicle (MAV) using visual snapshots is considered. The current problem is part of a larger project, which is developing an autonomous MAV, controlled by vision only information. A new algorithm is proposed that uses a stored image of the ground, a snapshot taken of the ground directly under the MAV, as a visual anchor point. The absolute translation of the aircraft and its velocity are then calculated by comparing the subsequent frames with the stored image and fed into the position controller. In order to increase the performance, several issues, such as effects of scale uncertainty on the closed loop stability of the platform are investigated. For controller design and testing purposes, we analytically derive a complete model of a small size helicopter with no stabilizing bar (flybar). The simulation results for 2D and 3D snapshots confirm the effectiveness of the proposed algorithm.

Put Your Best Figure Forward: Line Graphs and Scattergrams

There is an old saying that “a picture is worth a thousand words.” In truth, only a well-prepared, self-explanatory picture is worth a thousand words. The same holds true for research studies, for which 1 of the main methods we use to communicate our message is in figures and graphs. Figures and graphs tell much of the story by giving readers a visual anchor to help them see, understand, and remember information. Think about a report that you recently read and found useful. You likely do not remember the text used to state the results, or even the actual numbers, but you can recall much about the trends, relationships, outcomes, categories, or general experimental parameters shown in a graph. Despite the fact that you no longer have much recollection of the text, you can draw a reasonable representation of a graph from the published report and tell what you remember from it. In this educational article I discuss line graphs and scattergrams and use examples to illustrate how to put your best figure forward so readers will remember you and your message. The components of a graph include axes, labels, scales, an origin, tick or reference marks, symbols, and a legend. Beyond these basics, however, a good graph has several attributes: 1. It draws attention to the data and not the graph itself. 2. The data points (symbols) and connecting lines are easy to read and distinguish. 3. Both the numbers and labels for the axes are readable and their meaning is clear. 4. The lengths of the 2 axes are visually balanced (ratio of x axis to y axis = 1.0 to 1.3). 5. The scales used on each axis match the range of the data. 6. Tick marks are used appropriately. 7. The legend is clear and concise. 8. The reader can understand the message without …

Temporal facilitation of gaze in the presence of postural reactions triggered by sudden surface perturbations

Saccadic reaction times can be shortened by an additional sensory modality (e.g. auditory, tactile) presented in temporal proximity to the triggering cue. Whereas somatosensory cues given by sudden perturbations of the support surface can trigger appropriate postural adjustments to maintain upright stance, it is not known how gaze executions are affected by the dual task of maintaining upright balance while redirecting gaze. It was hypothesized that the onset latency of gaze movements toward visual targets will be shortened by sudden surface perturbations following visual target shifts to prompt a stable visual anchor for postural stabilization. Eight subjects stood on a movable platform with gaze fixated on a central target 2 m directly in front, and were instructed to shift their gaze to lateral targets located along a 63° arc to the right and left. The trials began with the central target lit followed randomly by either the right, left or center target. Fifty or 250 ms following this target shift, balance was perturbed by a sudden yaw movement of the support surface (15.5° over 210 ms at 130°/s), with no stepping or large arm reactions observed. The latency of the gaze shifts was significantly shortened (by ∼72 ms) when executed simultaneously with a surface perturbation. A decrease in excitation latency was also observed in the cervical paraspinals and sternocleidomastoid muscles. Postural responses in the ankle and knee muscles were not affected by gaze shifts. Pelvic horizontal angular motion closely followed surface motion whereas head motion was influenced by gaze shifts. During the combined gaze shift and surface motion conditions, thorax movement excursion was larger and not correlated with either the surface motion or visual target shift. In conclusion, postural adjustments in response to sudden surface yaws facilitate voluntary gaze shift execution and this enhancement may result from the sensory fusion of somatosensory and visual information.

Visual Impact Assessment of Landmark by rating Orientation

The concept of visual anchor is introduced to explain the spatial orientation and rating its degree. The spatial orientation contrib utes to urban legibility and it is important to control by methods in urban planning. 26 subjects evaluate the effect of perspective composition including landmark on the orientation. They are requested to point out from where some pictures of landmark are taken on the map. The degree of orientation is rated as errors on distance, location, and direction between standing point and landmark. Results imply that the spatial relationship of visual elements on the perspective image is effective.

Aging and postural control: postural perturbations caused by changing the visual anchor.

To determine the effect of modifying the stable visual anchor on the postural stability of older individuals. The visual anchor was changed by opening doors similar to those found in an elevator cage. Lighting intensities inside and outside the cage were varied to create increasing or decreasing luminosity conditions. The effect of adding a cognitive load (counting backwards) was also tested. A controlled laboratory study. Tests performed in a balance laboratory. Sensory and clinical measurements to insure the integrity of the central and peripheral nervous system. Measures of balance were derived from the recordings of the center of foot pressure. These measures included range and speed of the center of foot pressure. Eight older, community-dwelling subjects and nine young subjects participated. A sensorimotor evaluation was used to insure that all older individuals were free from any pathologies affecting postural stability. All participants had a low score (indicating high balance confidence) on the Falls Efficacy Scale and no history of falls. Older individuals were affected by modification of the stable visual anchor induced by the opening of doors similar to that of an elevator cage. They showed greater ranges of the center of foot pressure (COP) and speed of the COP after than before the opening of the doors. Furthermore, the increased ranges and speed were two to three times greater than that observed for the young subjects. A lighting intensity considered as comfortable for reading inside the elevator affected the overall postural stability of the older participants negatively. Counting backwards also decreased their overall stability. Changing the stable visual anchor, as when exiting an elevator cage, could be a significant risk factor for older persons. Moreover, when combined with a cognitive load or lower lighting intensity inside the elevator cage, the negative effects on the postural stability of older persons are exacerbated.

Sharing Teaching Ideas: Table Tennis, Anyone? Using Ping-Pong to Teach the Coordinate Plane

During a first-year-algebra class, tenth- and eleventh-grade students, overhearing the sounds of a Ping-Pong game from downstairs, voiced a strong desire to play. This incident led the authors to stumble onto the idea of using Ping-Pong to teach coordinate geometry. From this inquiry by interested students came the idea of using a Ping-Pong table as the visual anchor on which to situate the coordinate plane.