From ImageNet to mining: Adapting visual object detection with minimal supervision

This paper presents visual detection and classification of light vehicles and personnel on a mine site.We capitalise on the rapid advances of ConvNet based object recognition but highlight that a naive black box approach results in a significant number of false positives. In particular, the lack of domain specific training data and the unique landscape in a mine site causes a high rate of errors. We exploit the abundance of background-only images to train a k-means classifier to complement the ConvNet. Furthermore, localisation of objects of interest and a reduction in computation is enabled through region proposals. Our system is tested on over 10km of real mine site data and we were able to detect both light vehicles and personnel. We show that the introduction of our background model can reduce the false positive rate by an order of magnitude.

The impact of simulated astigmatism on functional measures of visual performance

Uncorrected refractive error, including astigmatism, is a leading cause of reversible visual impairment. While the ability to perform vision-related daily activities is reduced when people are not optimally corrected, only limited research has investigated the impact of uncorrected astigmatism. Given the capacity to perform vision-related daily activities involves integration of a range of visual and cognitive cues, this research examined the impact of simulated astigmatism on visual tasks that also involved cognitive input. The research also examined whether the higher levels of complexity inherent in Chinese characters makes them more susceptible to the effects of astigmatism. The effects of different powers of astigmatism, as well as astigmatism at different axes were investigated in order to determine the minimum level of astigmatism that resulted in a decrement in visual performance.

Enhanced student learning in accounting utilising web-based technology, peer-review feedback and reflective practices: A learning community approach to assessment

Higher education is becoming a major driver of economic competitiveness in an increasingly knowledge-driven global economy. Maintaining the competitive edge has seen an increase in public accountability of higher education institutions through the mechanism of ranking universities based on the quality of their teaching and learning outcomes. As a result, assessment processes are under scrutiny, creating tensions between standardisation and measurability and the development of creative and reflective learners. These tensions are further highlighted in the context of large undergraduate subjects, learner diversity and time-poor academics and students. Research suggests that high level and complex learning is best developed when assessment, combined with effective feedback practices, involves students as partners in these processes. This article reports on a four-phase, cross-institution and cross-discipline project designed to embed peer-review processes as part of the assessment in two large, undergraduate accounting classes. Using a social constructivist view of learning, which emphasises the role of both teacher and learner in the development of complex cognitive understandings, we undertook an iterative process of peer review. Successive phases built upon students’ feedback and achievements and input from language/learning and curriculum experts to improve the teaching and learning outcomes.

An upstream open reading frame is essential for feedback regulation of ascorbate biosynthesis in arabidopsis

Ascorbate (vitamin C) is an essential antioxidant and enzyme cofactor in both plants and animals. Ascorbate concentration is tightly regulated in plants, partly to respond to stress. Here, we demonstrate that ascorbate concentrations are determined via the posttranscriptional repression of GDP-l-galactose phosphorylase (GGP), a major control enzyme in the ascorbate biosynthesis pathway. This regulation requires a cis-acting upstream open reading frame (uORF) that represses the translation of the downstream GGP open reading frame under high ascorbate concentration. Disruption of this uORF stops the ascorbate feedback regulation of translation and results in increased ascorbate concentrations in leaves. The uORF is predicted to initiate at a noncanonical codon (ACG rather than AUG) and encode a 60- to 65-residue peptide. Analysis of ribosome protection data from Arabidopsis thaliana showed colocation of high levels of ribosomes with both the uORF and the main coding sequence of GGP. Together, our data indicate that the noncanonical uORF is translated and encodes a peptide that functions in the ascorbate inhibition of translation. This posttranslational regulation of ascorbate is likely an ancient mechanism of control as the uORF is conserved in GGP genes from mosses to angiosperms.

Local Commons: A visual approach to collective city making through situated community engagement

Due to the numerous possibilities of voicing concerns and the flood of data we are exposed to, local issues are sometimes at risk of being overlooked. This study explores Local Commons, a design intervention in public space that combines situated digital and tangible media in order to engage communities in contributing and debating different perspectives on a given local issue. The intervention invited the community to submit images of their perspectives on the issue, which were displayed on a public screen. Via tangible buttons in front of the screen, community members then agree or disagree on the displayed perspectives, creating a space for deliberation. In a user study, we were specifically interested in testing three aspects of our intervention, which are discussed in this paper: The difference that situatedness, visual content, and tangible interaction can make to urban community engagement.

The subcortical confinement hypothesis: A neurological model for schizotypal hallucinations

Introduction Different types of hallucinations are symptomatic of different conditions. Schizotypal hallucinations are unique in that they follow existing delusional narrative patterns: they are often bizarre, they are generally multimodal, and they are particularly vivid (the experience of a newsreader abusing you personally over the TV is both visual and aural. Patients who feel and hear silicone chips under their skin suffer from haptic hallucinations as well as aural ones, etc.) Although there are a number of hypotheses for hallucinations, few cogently grapple the sheer bizarreness of the ones experienced in schizotypal psychosis. Methods A review-based hypothesis, traversing theory from the molecular level to phenomenological expression as a distinct and recognizable symptomatology. Conclusion Hallucinations appear to be caused by a two-fold dysfunction in the mesofrontal dopamine pathway, which is considered here to mediate attention of different types: in the anterior medial frontal lobe, the receptors (largely D1 type) mediate declarative awareness, whereas the receptors in the striatum (largely D2 type) mediate latent awareness of known schemata. In healthy perception, most of the perceptual load is performed by the latter: by the top-down predictive and mimetic engine, with the bottom-up mechanism being used as a secondary tool to bring conscious deliberation to stimuli that fails to match up against expectations. In schizophrenia, the predictive mode is over-stimulated, while the bottom-up feedback mechanism atrophies. The dysfunctional distribution pattern effectively confines dopamine activity to the striatum, thereby stimulating the structural components of thought and behaviour: well-learned routines, narrative structures, lexica, grammar, schemata, archetypes, and other procedural resources. Meanwhile, the loss of activity in the frontal complex reduces the capacity for declarative awareness and for processing anything that fails to meet expectations.

Visual demands of modern primary classrooms and the impact of refractive anomalies on academic performance

This research investigated the visual demands in modern primary school classrooms and also the impact of common refractive anomalies on a child's ability to perform academic-related tasks. The results showed that relatively high levels of visual acuity, contrast demand and sustained accommodative-convergence are required to perform optimally in the modern classroom environment. It was also demonstrated that relatively low magnitudes of uncorrected refractive error may have a detrimental impact on children's ability to perform academic-related activities at school, with sustained near work further exacerbating this effect. These findings have important implications for both eye care practitioners and education authorities.

Translating Management Storylines: developing a visual strategy to communicate science to stakeholder community, a case for Mackay, QLD

This project developed a visual strategy and graphic outcomes to communicate the results of a scientific collaborative project to the Mackay community. During 2013 and 2014 a team from CSIRO engaged with the community in Mackay to collaboratively develop a set of strategies to improve the management of the Great Barrier Reef. The result of this work was a 300+ page scientific report that needed to be translated and summarised to the general community. The aim of this project was to strategically synthesise information contained in the report and to design and produce an outcome to be distributed to the participant community. By working with the CISRO researchers, an action toolkit was developed, with twelve cards and a booklet. Each card represented the story behind a certain local management issue and the actions that the participants suggested should be taken in order to improve management of The Reef. During the design synthesis it was identified that for all management issues there was a reference to the need to develop some sort of "educational campaign" to the area. That was then translated as an underlying action to support all other actions proposed in the toolkit.

An Intelligent Driver Warning System

This chapter is focussed on the research and development of an intelligent driver warning system (IDWS) as a means to improve road safety and driving comfort. Two independent IDWS case studies are presented. The first study examines the methodology and implementation for attentive visual tracking and trajectory estimation for dynamic scene segmentation problems. In the second case study, the concept of driver modelling is evaluated which can be used to provide useful feedback to drivers. In both case studies, the quality of IDWS is largely determined by the modelling capability for estimating multiple vehicle trajectories and modelling driving behaviour. A class of modelling techniques based on neural-fuzzy systems, which exhibits provable learning and modelling capability, is proposed. For complex modelling problems where the curse of dimensionality becomes an issue, a network construction algorithm based on Adaptive Spline Modelling of Observation Data (ASMOD) is also proposed.

Visual control of an electrohydraulic manipulator for underground mining applications

A large range of underground mining equipment makes use of compliant hydraulic arms for tasks such as rock-bolting, rock breaking, explosive charging and shotcreting. This paper describes a laboratory model electo-hydraulic manipulator which is used to prototype novel control and sensing techniques. The research is aimed at improving the safety and productivity of these mining tasks through automation, in particular the application of closed-loop visual positioning of the machine's end-effector.

Complete-linkage clustering for voice activity detection in audio and visual speech

We propose a novel technique for conducting robust voice activity detection (VAD) in high-noise recordings. We use Gaussian mixture modeling (GMM) to train two generic models; speech and non-speech. We then score smaller segments of a given (unseen) recording against each of these GMMs to obtain two respective likelihood scores for each segment. These scores are used to compute a dissimilarity measure between pairs of segments and to carry out complete-linkage clustering of the segments into speech and non-speech clusters. We compare the accuracy of our method against state-of-the-art and standardised VAD techniques to demonstrate an absolute improvement of 15% in half-total error rate (HTER) over the best performing baseline system and across the QUT-NOISE-TIMIT database. We then apply our approach to the Audio-Visual Database of American English (AVDBAE) to demonstrate the performance of our algorithm in using visual, audio-visual or a proposed fusion of these features.

Selectivity of human retinotopic visual cortex to S-cone-opponent, L/M-cone-opponent and achromatic stimulation

Our aim was to make a quantitative comparison of the response of the different visual cortical areas to selective stimulation of the two different cone-opponent pathways [long- and medium-wavelength (L/M)- and short-wavelength (S)-cone-opponent] and the achromatic pathway under equivalent conditions. The appropriate stimulus-contrast metric for the comparison of colour and achromatic sensitivity is unknown, however, and so a secondary aim was to investigate whether equivalent fMRI responses of each cortical area are predicted by stimulus contrast matched in multiples of detection threshold that approximately equates for visibility, or direct (cone) contrast matches in which psychophysical sensitivity is uncorrected. We found that the fMRI response across the two colour and achromatic pathways is not well predicted by threshold-scaled stimuli (perceptual visibility) but is better predicted by cone contrast, particularly for area V1. Our results show that the early visual areas (V1, V2, V3, VP and hV4) all have robust responses to colour. No area showed an overall colour preference, however, until anterior to V4 where we found a ventral occipital region that has a significant preference for chromatic stimuli, indicating a functional distinction from earlier areas. We found that all of these areas have a surprisingly strong response to S-cone stimuli, at least as great as the L/M response, suggesting a relative enhancement of the S-cone cortical signal. We also identified two areas (V3A and hMT+) with a significant preference for achromatic over chromatic stimuli, indicating a functional grouping into a dorsal pathway with a strong magnocellular input.

The dynamics of global visual effects and games development industries: lessons for Australia’s creative industries development policy

The paper critiques the focus of creative industries policy on capability development of small and medium sized firms and the provision of regional incentives. It analyses factors affecting the competitiveness and sustainability of the games development industry and visual effects suppliers to feature films. Interviews with participants in these industries highlight the need for policy instruments to take into consideration the structure and organization of global markets and the power of lead multinational corporations. We show that although forms of economic governance in these industries may allow sustainable value capture, they are interrupted by bottlenecks in which ferocious competition among suppliers is confronted by comparatively little competition among the lead firms. We argue that current approaches to creative industries policy aimed at building self-sustaining creative industries are unlikely to be sufficient because of the globalized nature of the industries. Rather, we argue that a more profitable approach is likely to require supporting diversification of the industries as ‘feeders’ into other areas of the economy.

Acoustic adaptation in cross database audio visual SHMM training for phonetic spoken term detection

Visual information in the form of lip movements of the speaker has been shown to improve the performance of speech recognition and search applications. In our previous work, we proposed cross database training of synchronous hidden Markov models (SHMMs) to make use of external large and publicly available audio databases in addition to the relatively small given audio visual database. In this work, the cross database training approach is improved by performing an additional audio adaptation step, which enables audio visual SHMMs to benefit from audio observations of the external audio models before adding visual modality to them. The proposed approach outperforms the baseline cross database training approach in clean and noisy environments in terms of phone recognition accuracy as well as spoken term detection (STD) accuracy.

Incorporating visual information for spoken term detection

Spoken term detection (STD) is the task of looking up a spoken term in a large volume of speech segments. In order to provide fast search, speech segments are first indexed into an intermediate representation using speech recognition engines which provide multiple hypotheses for each speech segment. Approximate matching techniques are usually applied at the search stage to compensate the poor performance of automatic speech recognition engines during indexing. Recently, using visual information in addition to audio information has been shown to improve phone recognition performance, particularly in noisy environments. In this paper, we will make use of visual information in the form of lip movements of the speaker in indexing stage and will investigate its effect on STD performance. Particularly, we will investigate if gains in phone recognition accuracy will carry through the approximate matching stage to provide similar gains in the final audio-visual STD system over a traditional audio only approach. We will also investigate the effect of using visual information on STD performance in different noise environments.