Vision Australia

Tactile graphics are widely used to present maps and statistical diagrams to blind and low vision (BLV) people, with accessibility guidelines recommending their use for graphics where spatial relationships are important. Their use is expected to grow with the advent of commodity refreshable tactile displays. However, in stark contrast to visual information graphics, we lack a clear understanding of the benefts that well-designed tactile information graphics offer over text descriptions for BLV people. To address this gap, we introduce a framework considering the three components of encoding, perception and cognition to examine the known benefts for visual information graphics and explore their applicability to tactile information graphics. This work establishes a preliminary theoretical foundation for the tactile-frst design of information graphics and identifes future research avenues.

To develop a model investigating the effects of the psychometric function's slope and lapse rate on QUEST visual acuity (VA) thresholds for preschool children. Monte Carlo simulations of QUEST (16 trials) measurements of VA were performed using an underlying ideal observer model for a range of Ideal Thresholds, Slopes, and Lapse Rates for 10- and 4-alternative forced choices (AFC). Ideal Slopes (in probit size) were varied from steep to flat, 0.05 to 0.20 logMAR. Ideal Lapse Rates were varied from 0.01 to 0.50 (at Ideal Probit Size of 0.10 logMAR). Thresholds were estimated using three different techniques: (i) QUEST's maximum likelihood estimate (75% correct), and probit analysis of logMAR value (55% correct) with (ii) fixed-slope probit (probit size=0.10 logMAR) or (iii) floating-slope probit (probit size=0.02-0.29 logMAR). An index, Gain-Corrected Standard Deviation, was developed to assess how sensitive a technique was for detecting a change in VA. A smaller Gain-Corrected Standard Deviation indicates a technique is more sensitive to change. Measured VA thresholds from QUEST, floating-slope probit, and fixed-slope probit are linearly related to Ideal Threshold from -0.20 to 0.80 logMAR. Flat slopes and higher lapse rates result in an increased Gain-Corrected Standard Deviation for all threshold measurement techniques (higher when thresholds were estimated with probit fitting than with QUEST Maximum likelihood), with the increase in Gain-Corrected Standard Deviation being slightly higher for 4AFC compared to 10AFC across the range of Ideal Slopes and Lapse Rates. For 4AFC, the Gain-Corrected Standard Deviation range for varied lapse rates (0.01-0.50) was QUEST, 0.062-0.567; floating-slope probit, 0.079-0.623; and fixed-slope probit, 0.079-0.626 logMAR. QUEST maximum likelihood outputs should be used instead of fitting probit functions in preschool children when there are limited numbers of trials. Flat slopes and high lapse rates can significantly decrease the precision of QUEST estimates of VA in preschool children.

Respiratory syncytial virus (RSV) is a major cause of respiratory infection with a higher burden in Aboriginal and Torres Strait Islander infants and children. We conducted a pilot qualitative study identifying disease knowledge and willingness to immunise following the changing immunisation landscape for infant RSV in 2024. Yarning groups were held with a convenience sample of parents/carers of Aboriginal children attending playgroup at a metropolitan Aboriginal Health Service in Western Australia. Data collected in the form of notes were thematically analysed. We heard from nine parents/carers over two yarns in March/April 2024. Level of RSV awareness largely depended on lived experience of an RSV infection with some participants only first hearing of RSV following announcement of the immunisation program. Most participants were willing to accept immunisation. There was a strong preference for information on disease and immunisation safety coming from a 'trusted Aboriginal voice', but the level of information varied. This pilot study provides initial insights into community views of RSV disease and immunisation. More RSV disease awareness is needed in the Indigenous community. These findings will help inform current and future RSV immunisation programs for Aboriginal and Torres Strait Islander populations.

ABSTRACT How does a social cause like race in public education get revived and reinvented? How do political interest groups mount a campaign that flows internationally from the United States to Australia and leads to the political exclusion of Critical Race Theory from the Australian Curriculum? This article addresses these questions by examining how social media has been used to distribute propaganda that infers racial literacy education is a problem for Australian schools. Drawing on Hannah Arendt’s theories of propaganda and using artefacts from multiple Internet websites, including social media, this article examines how conservative lobbyists in education have adjusted claims from the United States to suit the highly mediated political culture of Australia in 2021. These conservative groups worked to gain political ground by deliberately subverting Critical Race Theory, turning it into a political weapon that has been deployed by political actors across the US, UK and Australia. This article traces the subversion of CRT and efforts by social media creators, editors and academics to control the meaning of the Law School concept. It concludes by discussing the consequences the use of CRT as a political weapon has had on Australian education.

High-quality training data is essential for enhancing the robustness of object detection models. Within the maritime domain, obtaining a diverse real image dataset is particularly challenging due to the difficulty of capturing sea images with the presence of maritime objects, especially in stormy conditions. These challenges arise due to resource limitations, in addition to the unpredictable appearance of maritime objects. Nevertheless, acquiring data from stormy conditions is essential for training effective maritime detection models, particularly for search and rescue, where real-world conditions can be unpredictable. In this work, we introduce SafeSea, which is a stepping stone towards transforming actual sea images with various Sea State backgrounds while retaining maritime objects. Compared to existing generative methods such as Stable Diffusion Inpainting [27], this approach reduces the time and effort required to create synthetic datasets for training maritime object detection models. The proposed method uses two automated filters to only pass generated images that meet the criteria. In particular, these filters will first classify the sea condition according to its Sea State level and then it will check whether the objects from the input image are still preserved. This method enabled the creation of the SafeSea dataset, offering diverse weather condition backgrounds to supplement the training of maritime models. Lastly, we observed that a maritime object detection model faced challenges in detecting objects in stormy sea backgrounds, emphasizing the impact of weather conditions on detection accuracy. The code, and dataset are available at https://github.com/martin-3240/SafeSea.

There has been a notable advancement in the task of object detection over the last few years. However, recent studies reveal that modern object detectors suffer a considerable performance drop in an open-set setting and tend to incorrectly detect unknown objects and mislabel them as belonging to known categories with a high degree of confidence. We observe that the ‘catch-all background’ approach with limited training data from the detection dataset is insufficient to equip a detector with the ability to reject all unknown objects as background. Based on this observation, we propose to augment the detection dataset by reusing a subset of ImageNet that does not contain any object categories predefined by the detection task. The augmented training on a held-in background dataset allows the detector to reject unknown object instances as background instead of mislabelling them. Our experiments demonstrate that the proposed simple yet effective REusing ImageNet (REIN) technique improves the performance of object detectors in an open-set setting. Given the promising results, we hope that our method can encourage our community to pay more attention to open-set object detection and to rethink the data augmentation in detection.

Digitisation of physical insect collections in the world is essential to understanding the biodiversity, and creating digital duplicates in 3D that can be easily shared with scientists and public users. This paper presents a fully automatic system for fast digitising large insect collections. The focus is on automating both insect handling and insect scanning to perform whole-drawer 3D reconstruction of all specimens in an insect tray. Though existing approaches have worked on automating insect scanning, those methods often overlook automatic insect handling. Current insect handling generally involves trained manual labour and is time-consuming. To resolve this issue, we propose to use a robot arm that can perform the insect handling efficiently with high repeatability. In our approach, insects are automatically localised, grasped, scanned, and reconstructed in 3D using the cameras with the robot arm. This results in scalable high-throughput automated system that is capable of handling multiple trays of insects. Our system is generalizable and can be used to digitise other types of physical collections by just replacing the collection-specific object detector in our pipeline.

Hyperopia is associated with reduced vision and educational outcomes in schoolchildren. This study explored the impact of clinically significant hyperopia (≥+2.00 D) on visual function in schoolchildren and compared the ability of different screening tests (alone and in combination) to detect this level of hyperopia. Vision testing including monocular logMAR visual acuity (VA) measured to threshold (distance [DVA], near [NVA] and DVA through a plus lens [+2.50 D]), stereoacuity and cycloplegic autorefraction (tropicamide 1%) were undertaken on 263 schoolchildren (mean age: 11.76 years ± 3.38) in Queensland, Australia. Vision measures were compared between children with clinically significant hyperopia in at least one meridian (≥+2.00 D) and emmetropia/low hyperopia (>0.00 and <+2.00 D). Receiver operating curve (ROC) analysis was performed to identify optimal pass/fail criteria for each test and the diagnostic accuracy of individual and combinations of tests. Thirty-two children had clinically significant hyperopia and 225 had emmetropia/low hyperopia. DVA and NVA were worse (p < 0.01), while the difference in DVA through a plus lens was less in children with clinically significant hyperopia (p < 0.01). ROC analysis for individual tests resulted in areas under the curve (AUCs) ranging from 0.65 to 0.85. Combining screening tests revealed that failing one or more of the following tests was most effective for detecting hyperopia: DVA, NVA and difference in DVA through a plus lens, resulting in a sensitivity and specificity of 72% and 81%, respectively. Significant differences in visual function existed between schoolchildren with clinically significant hyperopia and emmetropia/low hyperopia. Combining measures of DVA and NVA and the difference in DVA through a plus lens demonstrated good discriminative ability for detecting clinically significant hyperopia in this population.

BackgroundAs the demand for youth mental health care continues to rise, managing wait times and reducing treatment delays are key challenges to delivering timely and quality care. Clinical staging is a heuristic model for youth mental health that can stratify care allocation according to individuals’ risk of illness progression. The application of staging has been traditionally limited to trained clinicians yet leveraging digital technologies to apply clinical staging could increase the scalability and usability of this model in services.ObjectiveThe aim of this study was to validate a digital algorithm to accurately differentiate young people at lower and higher risk of developing mental disorders.MethodsWe conducted a study with a cohort comprising 131 young people, aged between 16 and 25 years, who presented to youth mental health services in Australia between November 2018 and March 2021. Expert psychiatrists independently assigned clinical stages (either stage 1a or stage 1b+), which were then compared to the digital algorithm’s allocation based on a multidimensional self-report questionnaire.ResultsOf the 131 participants, the mean age was 20.3 (SD 2.4) years, and 72% (94/131) of them were female. Ninety-one percent of clinical stage ratings were concordant between the digital algorithm and the experts’ ratings, with a substantial interrater agreement (κ=0.67; P<.001). The algorithm demonstrated an accuracy of 91% (95% CI 86%-95%; P=.03), a sensitivity of 80%, a specificity of 93%, and an F1-score of 73%. Of the concordant ratings, 16 young people were allocated to stage 1a, while 103 were assigned to stage 1b+. Among the 12 discordant cases, the digital algorithm allocated a lower stage (stage 1a) to 8 participants compared to the experts. These individuals had significantly milder symptoms of depression (P<.001) and anxiety (P<.001) compared to those with concordant stage 1b+ ratings.ConclusionsThis novel digital algorithm is sufficiently robust to be used as an adjunctive decision support tool to stratify care and assist with demand management in youth mental health services. This work could transform care pathways and expedite care allocation for those in the early stages of common anxiety and depressive disorders. Between 11% and 27% of young people seeking care may benefit from low-intensity, self-directed, or brief interventions. Findings from this study suggest the possibility of redirecting clinical capacity to focus on individuals in stage 1b+ for further assessment and intervention.

Vision standards for driving are typically based on visual acuity, despite evidence that it is a poor predictor of driving safety and performance. However, visual motion perception is potentially relevant for driving, as the vehicle and surroundings are in motion. This study explored whether tests of central and mid-peripheral motion perception better predict performance on a hazard perception test (HPT), which is related to driving performance and crash risk, than visual acuity. Additionally, we explored whether age influences these associations, as healthy ageing impairs performance on some motion sensitivity tests. Sixty-five visually healthy drivers (35 younger, mean age: 25.5; SD 4.3 years; 30 older adults, mean age: 71.0; SD 5.4 years) underwent a computer-based HPT, plus four different motion sensitivity tests both centrally and at 15° eccentricity. Motion tests included minimum displacement to identify motion direction (Dmin ), contrast detection threshold for a drifting Gabor (motion contrast), coherence threshold for a translational global motion stimulus and direction discrimination for a biological motion stimulus in the presence of noise. Overall, HPT reaction times were not significantly different between age groups (p = 0.40) nor were maximum HPT reaction times (p = 0.34). HPT response time was associated with motion contrast and Dmin centrally (r = 0.30, p = 0.02 and r = 0.28, p = 0.02, respectively) and with Dmin peripherally (r = 0.34, p = 0.005); these associations were not affected by age group. There was no significant association between binocular visual acuity and HPT response times (r = 0.02, p = 0.29). Some measures of motion sensitivity in central and mid-peripheral vision were associated with HPT response times, whereas binocular visual acuity was not. Peripheral testing did not show an advantage over central testing for visually healthy older drivers. Our findings add to the growing body of evidence that the ability to detect small motion changes may have potential to identify unsafe road users.