Search

Estimating depth from RGB images is a long-standing ill-posed problem, which has been explored for decades by the computer vision, graphics, and machine learning communities. Among the existing techniques, stereo matching remains one of the most widely used in the literature due to its strong connection to the human binocular system. Traditionally, stereo-based depth estimation has been addressed through matching hand-crafted features across multiple images. Despite the extensive amount of research, these traditional techniques still suffer in the presence of highly textured areas, large uniform regions, and occlusions. Motivated by their growing success in solving various 2D and 3D vision problems, deep learning for stereo-based depth estimation has attracted a growing interest from the community, with more than 150 papers published in this area between 2014 and 2019. This new generation of methods has demonstrated a significant leap in performance, enabling applications such as autonomous driving and augmented reality. In this paper, we provide a comprehensive survey of this new and continuously growing field of research, summarize the most commonly used pipelines, and discuss their benefits and limitations. In retrospect of what has been achieved so far, we also conjecture what the future may hold for deep learning-based stereo for depth estimation research. [ABSTRACT FROM AUTHOR]

Fewer studies have investigated the factors affecting fully Automated Vehicles (AVs) acceptance and their association with demographics and travel behaviour in the Australian context. Evidence shows that public opinions and adoption towards AVs may vary by country. Therefore, this study aims to systematically explore the factors affecting public opinion towards fully AVs by conducting a survey in Australia. Specifically, the study examines seven key attributes (benefits/usefulness, ease of use, attitude, data privacy, willingness to pay more, social influence, and trust) towards the opinions of AVs and their association between demographic and travel behaviour. The results from 809 Australian responses show that Australians tend to have a positive attitude toward AVs. Age and years of driving experience are the two most important factors instead of travel characteristics (e.g. frequency of driving cars, travel time by car, and distance of driving cars). Being less exposed to automation technology is likely to be the main reason affecting the adoption of AVs for the two socio‐demographic factors: age and years of driving experience. Males hold a relatively positive attitude towards AVs than females. Likewise, Australians are less likely to pay more for fully AVs than conventional vehicles. Data privacy is also concerning for Australians. [ABSTRACT FROM AUTHOR]

The capability of 'demand-responsive transport', particularly in autonomous shared form, to better facilitate road-based mobility is considered a significant advantage because improved mobility leads to enhanced quality of life and wellbeing. A central point in implementing a demand-responsive transit system in a new area is adapting the operational concept to the respective structural and socioeconomic conditions. This requires an extensive analysis of the users' needs. There is presently limited understanding of public perceptions and attitudes toward the adoption of autonomous demand-responsive transport. To address this gap, a theory-based conceptual framework is proposed to provide detailed empirical insights into the public's adoption intention of 'autonomous shuttle buses' as a form of autonomous demand-responsive transport. South East Queensland, Australia, was selected as the testbed. In this case study, relationships between perceptions, attitudes, and usage intention were examined by employing a partial least squares structural equation modeling method. The results support the basic technology acceptance model casual relationships that correspond with previous studies. Although the direct effects of perceived relative advantages and perceived service quality on usage intention are not significant, they could still affect usage intention indirectly through the attitude factor. Conversely, perceived risks are shown to have no association with perceived usefulness but can negatively impact travelers' attitudes and usage intention toward autonomous shuttle buses. The research findings provide implications to assist policymakers, transport planners, and engineers in their policy decisions and system plans as well as achieving higher public acknowledgment and wider uptake of autonomous demand-responsive transport technology solutions. [ABSTRACT FROM AUTHOR]

This paper describes insights gained from a decade of autonomous marine systems development at the University of Sydney’s Australian Centre for Marine Robotics. Over the course of this time, we have deployed numerous vehicles and imaging platforms in support of applications in engineering science, marine ecology, archaeology and geoscience. We have operated an Australia-wide benthic observing program designed to deliver precisely navigated, repeat imagery of the seafloor. This initiative makes extensive use of Autonomous Underwater Vehicles (AUVs) to collect high-resolution stereo imagery, multibeam sonar and water column measurements on an annual or semi-annual basis at sites around Australia, spanning the full latitudinal range of the continent from tropical reefs in the north to temperate regions in the south. We have also contributed to expeditions to document coral bleaching, cyclone recovery, submerged neolithic settlement sites, ancient shipwrecks, methane seeps and deepwater hydrothermal vents. We briefly consider how automated tools for working with this imagery have facilitated the resulting science outcomes. [ABSTRACT FROM AUTHOR]

This article explores the emergence of Automated Vehicles (AVs) in Australia. It will investigate the legal and regulatory terrain. International and domestic approaches are examined to determine potential responses. The regulatory issues emerge partly due to the varied nature of artificial intelligence systems and processes that enable AVs to function. The variations may be due to the chosen domain model, software development processes, or the development of biases that may occur during code development for the underlying artificial intelligence system. Such variations can create difficulties in the application of road rules, safety requirements, and the legal and regulatory requirements. They may give rise to significant issues relating to driver classification and licensing for AVs, due to the varied levels of control and involvement in the driving process. For this reason, legislative reform at specific jurisdictional levels is suggested together with clearer international standards as a pathway to ensure the safe and effective integration of autonomous vehicles into society. [ABSTRACT FROM AUTHOR]

Current carbon measurement strategies leave spatiotemporal gaps that hinder the scientific understanding of the oceanic carbon biogeochemical cycle. Data products and models are subject to bias because they rely on data that inadequately capture mesoscale spatiotemporal (kilometers and days to weeks) changes. Highresolution measurement strategies need to be implemented to adequately evaluate the global ocean carbon cycle. To augment the spatial and temporal coverage of ocean-atmosphere carbon measurements, an Autonomous Surface Vehicle CO2 (ASVCO2) system was developed. From 2011 to 2018, ASVCO2 systems were deployed on seven Wave Glider and Saildrone missions along the U.S. Pacific and Australia's Tasmanian coastlines and in the tropical Pacific Ocean to evaluate the viability of the sensors and their applicability to carbon cycle research. Here we illustrate that the ASVCO2 systems are capable of long-term oceanic deployment and robust collection of air and seawater pCO2 within ±2 µatm based on comparisons with established shipboard underway systems, with previously described Moored Autonomous pCO2 (MAPCO2) systems, and with companion ASVCO2 systems deployed side by side. [ABSTRACT FROM AUTHOR]

For the benefits of autonomous vehicles (AVs) to be optimized, the fleet conversion process needs to be efficient and timely. This study explored public attitudes to AVs to inform strategies to increase receptivity to the wide-scale use of AVs. A national online survey was administered to a sample of 1,624 Australians aged 16+ years. The survey featured open-ended questions that scoped respondents' perceptions of AVs. A grounded, thematic analysis identified two primary dimensions in the data: response valence (how positive or negative the comments were about the advent of AVs) and response type (the extent to which the comments reflected a cognitive or emotional response). This resulted in a dimensional analysis featuring four quadrants that captured the topics that were most frequently raised spontaneously by respondents. The quadrant characterized by comments that were positive/neutral and cognitive in nature was the most substantial, indicating general acceptance. Where concerns were expressed, they typically related to perceived safety, trust, and control issues, and tended to be more emotional in nature. The results highlight the importance of providing the public with concrete information about AVs to address fear levels and to resolve trust and control issues. [ABSTRACT FROM AUTHOR]

Assemblages of megabenthos are structured in seven depth-related zones between ∼700 and 4000 m on the rocky and topographically complex continental margin south of Tasmania, southeastern Australia. These patterns emerge from analysis of imagery and specimen collections taken from a suite of surveys using photographic and in situ sampling by epibenthic sleds, towed video cameras, an autonomous underwater vehicle and a remotely operated vehicle (ROV). Seamount peaks in shallow zones had relatively low biomass and low diversity assemblages, which may be in part natural and in part due to effects of bottom trawl fishing. Species richness was highest at intermediate depths (1000–1300 m) as a result of an extensive coral reef community based on the bioherm-forming scleractinian Solenosmilia variabilis. However, megabenthos abundance peaked in a deeper, low diversity assemblage at 2000–2500 m. The S. variabilis reef and the deep biomass zone were separated by an extensive dead, sub-fossil S. variabilis reef and a relatively low biomass stratum on volcanic rock roughly coincident with the oxygen minimum layer. Below 2400 m, megabenthos was increasingly sparse, though punctuated by occasional small pockets of relatively high diversity and biomass. Nonetheless, megabenthic organisms were observed in the vast majority of photographs on all seabed habitats and to the maximum depths observed - a sandy plain below 3950 m. Taxonomic studies in progress suggest that the observed depth zonation is based in part on changing species mixes with depth, but also an underlying commonality to much of the seamount and rocky substrate biota across all depths. Although the mechanisms supporting the extraordinarily high biomass in 2000–2500 m depths remains obscure, plausible explanations include equatorwards lateral transport of polar production and/or a response to depth-stratified oxygen availability. [ABSTRACT FROM AUTHOR]