Sydney eScholarship
Not a member yet
31878 research outputs found
Sort by
Carriers and Barriers of Voice in Teams: An Investigation into Team Voice and the Role of Team Voice Allies and Resistors in Influencing Voice Outcomes
No full textAs organisations increasingly rely on teams as foundational work units, understanding how voice unfolds within workgroups and teams has become a critical research topic. Moving beyond traditional, dyadic perspectives on employee voice, this thesis builds one merging perspectives to investigate team voice as a multifaceted and socially embedded process involving multiple actors. In this thesis, I present three empirical studies. In Study1, I conducted a comprehensive literature review and meta-analysis of 38 empirical studies on team voice, analysing key antecedents and consequences of team voice including both its promotive and prohibitive content. The results of this first study show that team voice is shaped by team leadership and team climate, and that team voice also influences team performance, team innovation and team viability. Study 2 uses an experimental design to investigate how team members, acting as voice allies or resistors, influence team voice and team wellbeing. The results of this second study showed that teams with voice allies showed higher voice, satisfaction, and positive affect, while teams with resistors showed lower voice and wellbeing. In Study 3, I investigated the intra-individual consequences of enacting voice allyship or resistance. The results of Study 3suggest that there are personal resource consequences to voice allies and resistors in enacting these roles. Overall, by integrating insights across multiple analytical levels and theoretical perspectives, I provide a richer understanding of how voice unfolds within modern, interdependent teams. Theoretically, I extend voice scholarship by highlighting the evolving roles of team members as collective voicers, voice allies, and resistors across both voice-related and emotional dimensions. Practically, I offer actionable insights into how organisations can foster sustainable voice by recognising and supporting the social dynamics that impact team voice and member wellbeing
Oral Health Research Connect - August 2025: Guest Speaker - Associate Professor James Tsoi
No full textTopic: 3D-PAIR (Printing/AI/Robotics) in Dentistry-Prof James Tsoi
Technologies are changing so fast, and the dental landscape is also leveraged and transformed from traditional to digital and then to autonomous dentistry. This lecture will be focusing on various (comparatively) new 3D-printing, AI and robotics (3D-PAIR) research and technologies from HKU Dental Materials Science. By utilizing 3D-PAIR as a tech-core and "mix & match" with new materials, designs and treatment modalities, we aim at creating new values for dentists, dental manufacturers, and patients/the general public
Final Report on the Alternative Formats and Support Options for the Disability Wellbeing Index (DWI): Promoting Accessibility, Safety, and Self Reporting for Survey Respondents
No full textThe Centre for Disability Research and Policy (CDRP), University of Sydney, was contracted to contribute to a three-year research project, 2022-2024, funded by the National Disability Insurance Agency (NDIA) and lead by Associate Professor Gang Chen, Monash University. The overarching aim of this research project was to design and test a preference-based wellbeing instrument that captures factors impacting on the wellbeing of people with disability in Australia, now known as the Disability Wellbeing Index (DWI). The role of the team based at the CDRP was to facilitate people with disabilities being involved in each stage of the research, including accessibility, safe environment, and self-reporting considerations for survey respondents.
This is the final of four reports documenting the contribution of the team at the CDRP. In this report, we describe the implementation of the alternative formats and support options offered to DWI survey respondents and our understanding of factors that facilitated (or did not) NDIS participants completing the DWI survey. These insights will help refine future iterations of the DWI and provide guidance for improving the provision of accessible communications to people with disability
Biophilic Design Experiences in the Workplace: An Immersive Virtual Reality Approach to Evaluating Restorative Benefits of Multisensory Stimuli
No full textPrevious research has extensively documented the benefits of interacting with nature. Biophilic design applications aimed at connecting the built environment back to nature have been shown to improve employee health, well-being, and productivity in workplaces. However, most restorative environmental studies have focused predominantly on visual qualities of nature, overlooking other senses. In particular, the role of smell in supporting restoration has been largely neglected. This thesis addresses this gap by exploring how olfactory stimuli contribute to the restorative experience of biophilic design in workplaces. It asks: “To what extent does the sense of smell contribute to the multisensory experience of biophilic design and its restorative benefits in workplace environments?” A pretest-posttest experiment was conducted comparing two distinct workplace conditions: one non-biophilic and another offering multisensory exposure to nature, including visual, olfactory, and auditory. A novel multisensory virtual reality system (MVRS) was developed to simulate the dynamic environmental conditions of an actual workplace in Sydney, Australia. Findings showed that the multisensory biophilic workplace significantly enhanced cognitive performance, reduced stress levels, and enhanced mood states compared to the non-biophilic one. There was a statistically significant association between the sense of smell and restorative benefits. While visual cues remained dominant, olfactory stimuli emerged as a key contributor to the overall restorative experience. This thesis provides rare empirical evidence on the importance of olfactory cues in nature-based experiences and restorative design, contributing to the knowledge within the realms of biophilia and challenging ocular-centrism approaches in architectural discourse. The methodological innovation and findings from the project invite further exploration of the restorative benefits of olfactory pleasure in wider contexts
DeCoP: Dependency Controlled Pre-training for Time Series Representation Learning
No full textMasked time series modeling (MTM) has become a leading approach in self-supervised pre-training for time series data. However, existing frameworks struggle to effectively model dependencies that balance informative signals and noise, often leading to overfitting or missing critical temporal dependencies due to non-stationarity and limited semantic context in time series data. To address this, we introduce DeCoP, a Dependency Controlled Pre-training framework that enhances self-supervised time series representation by effectively controlling dependency modeling, while significantly reducing computational cost. DeCoP controllably reduces non-stationary noise and disentangles mixed temporal variations from coarse to fine levels, thereby improving representation clarity. Specifically, DeCoP incorporates Instance-wise Patch Normalization (IPN) for controlled dependency, which reduces noise and stabilizes data distributions, establishing a controlled foundation for dependency modeling. Furthermore, a Hierarchical Dependency Controlled Learning (DCL) strategy is employed to selectively control inter-patch dependency ranges, generating robust, generalizable embeddings that enhance model stability across varying time series patterns. Extensive evaluations on ETTh1 datasets reveal that DeCoP achieves up to a 3% improvement in MSE over PatchTST, using only 37% of the FLOPs required
Near-term ecological forecasting: A Bayesian framework for modelling spatio-temporal population dynamics in extreme environments
No full textEcological forecasting plays a crucial role in anticipating ecosystem dynamics and guiding conservation decisions, particularly in arid ecosystems where species respond to extreme environmental variability and complex interactions. This thesis focuses on forecasting the population abundance of small mammals in central Australia, integrating long-term monitoring data with advanced statistical methods to improve model accuracy, ecological relevance, and practical use. Using a combination of Multivariate Autoregressive State-Space (MARSS) models and Multivariate Generalized Additive Models (MVGAM), the thesis explores how model structure, missing data, training length, and forecast horizon influence predictive performance. While MARSS models are effective in capturing structured population dynamics across space, they struggle with computational limitations and nonlinear processes. Simulated experiments evaluating imputation methods show that although differences in predictive accuracy are small, multiple imputation better preserves underlying ecological patterns and uncertainty. Forecasting performance is shown to depend more on model flexibility and the ability to represent latent ecological processes than on the quantity of training data alone. Short-term forecasts using projected climate data reveal divergent species responses across sites and climate scenarios, underscoring the importance of near-term ecological forecasting. Overall, this work demonstrates the value of combining ecological understanding with flexible, interpretable modelling approaches to improve forecasts in variable environments and inform conservation planning under climate change
Development of an advanced drug delivery system to prevent and treat breast cancer bone metastasis
No full textBreast cancer (BC) is the most common malignancy in women, with triple-negative BC (TNBC) comprising 15–20% of cases. TNBC lacks ER, PR, and HER2 receptors, making it unresponsive to targeted therapies; doxorubicin (Dox) remains the main treatment, though limited by toxicity and resistance. Curcumin (Cur) shows anticancer potential but suffers from low bioavailability. We developed a dual-drug nanocarrier co-delivering Cur and Dox, using alendronate (ALN) for bone targeting and LHRH for tumor specificity. Pluronic F127-based micelles and niosomes were optimized to enhance drug stability and delivery. F127 integration (Span 60:F127:Cholesterol, 0.99:0.01:1) improved encapsulation efficiency (EE%): Dox EE% rose to 58.6% and Cur to 98% in F127-containing niosomes (CDFN), compared to 18.3% and 30.1% in controls. FTIR and XPS confirmed successful ligand conjugation. DLS/SEM showed particle sizes of 20–32 nm for micelles and 244.6 nm for CDFN. Zeta potential shifted from –26.2 mV (F127) to –10.3 mV (PEGylation), indicating stability. [k⁶(Ahx)]-LHRH-Cur micelles showed the lowest IC₅₀ (3.2 µM), and CDFN had superior cytotoxicity (IC₅₀ = 0.7671 µM) vs. free Dox (5.48 µM) and Cur (124.6 µM), with a synergistic CI of 0.271. In 3D spheroids, CDFN caused near-complete disintegration by day 4. Calcium-binding assays confirmed ALN-mediated bone targeting (p < 0.0001). CDFN remained stable for 7 days at room temperature and released drugs faster at acidic pH. This dual-drug system shows enhanced efficacy and selectivity, offering a promising strategy for breast cancer and bone metastasis treatment
Neurophysiological Signals Analysis with Nanoelectronic System Platforms
No full textTo address the issue of high costs in physiological signal processing, this paper combines memristor technology with brain-inspired computing methods to explore a low-power, high-performance on-chip learning and inference system to improve the efficiency and accuracy of Electroencephalogram (EEG) and Electrocardiogram (ECG) signal processing.
This study proposes two novel neural network architectures, including the Kolmogorov-Arnold Network (KAN) based on learnable activation functions and the Spiking Neural Network (SNN) combined with dendrites based on neuromorphic devices. By introducing memristors as hardware implementations for neurons and synapses, an efficient memristor array was constructed, improving the ability to integrate algorithms with hardware systems. In addition, based on memristor technology, we designed an innovative neuron model based on dendritic Leaky Integrated and Fire (LIF) neurons that improve network performance without relying on the traditional delay layer approach. This not only enhances the spatiotemporal processing capability of the network for physiological signals but also improves the system's biological plausibility and computational efficiency.
Experimental results show that the proposed networks achieve strong performance in EEG seizure prediction and ECG abnormality detection. They offer effective physiological signal processing with significantly reduced power consumption, delivering comparable or superior accuracy at lower computational cost than conventional neural networks.
This work presents a novel solution for low-power physiological signal processing, supporting applications in intelligent medical devices and brain-computer interfaces. Future research will focus on optimising memristor dynamics, extending to more signal modalities, and facilitating large-scale hardware integration
An evolutionary-institutionalist framework for interpreting industry policy: the case of Australia, 1942-2022
No full textThis thesis develops a new interpretive framework for analyzing the evolution of industry policy, addressing the limitations of approaches based on neoclassical economics and international relations. Drawing on Institutional Economics — particularly the Regulation-Social Structure of Accumulation approach and evolutionary economics — it examines Australian industry policy across three key periods: post-World War II reconstruction (1942–73), structural adjustment and liberalisation (1974–98), and energy transformation and securitisation (1999–2022).
It identifies economic and political instabilities driving policy change, including balance of payments constraints, sectoral and regional misalignments, national security concerns, technology gaps, and inequality. It explores how these were managed through institutional mechanisms such as industry assistance bodies, arbitration tribunals, and trade and defence diplomacy. Industry policies are categorised into six strategies: access-attract, stabilise-redistribute, diversify-upgrade, defend-securitise, innovate-compete, and pioneer-structure.
The thesis highlights the path-dependent effects of earlier policies on later challenges. For example, tariff reliance in the mid-20th century complicated reforms in the 1970s–80s. Integration with North-East Asia in the 1990s introduced new vulnerabilities, while trade liberalisation and resource dependence in the 2000s created barriers to clean energy transition. Recent stagnation in income and economic complexity has driven renewed defend-securitise strategies.
Rejecting normative evaluations of policy efficiency, the study explains industry policy as a structural response to systemic instability. The Australian case shows how strategies evolve under distinct historical conditions, offering insights for comparative research in international political economy and industrial transformation
Learning Semantic Consistency for Robust Image Segmentation
No full textImage segmentation models often fail when they encounter unseen lighting, motion blur, or domain shift between synthetic and real data. This dissertation proposes a unified framework—Semantic‑Consistency Learning (SCL)—to deliver stable, fine‑grained segmentation under such challenging conditions. First, a dual‑branch architecture aligns class‑specific features from weakly and fully supervised streams, allowing the network to learn from sparse or noisy labels while keeping annotation costs low. Next, a temporal self‑ensembling strategy tracks an exponential moving average of network parameters, smoothing gradients and increasing resistance to distribution shift. Third, frequency‑domain perturbations and adaptive contrastive losses are introduced to suppress artefacts caused by weather, night scenes, and motion. Comprehensive tests on Cityscapes, ACDC, SynROD, and custom UAV datasets show mean‑IoU gains of 3–5 % over strong baselines, with a 40 % reduction in required pixel annotations. Ablation studies confirm that each consistency constraint contributes additively to accuracy and generalisation. The thesis also provides theoretical bounds explaining why multi‑scale consistency regularises feature space and improves robustness. Finally, an open‑source 3‑D visualisation toolkit is released, enabling interactive scrutiny of uncertainty maps and facilitating downstream use in autonomous driving and low‑altitude logistics planning. Taken together, these contributions advance the practical deployment of segmentation models in safety‑critical and resource‑constrained scenarios.
Suggested keywords (≤ 6, comma‑separated)
semantic segmentation, consistency learning, domain adaptation, deep learning, robust vision, annotation efficienc