47011 research outputs found
Sort by
Extraction of optical properties from scattering media using a convolutional neural network and diverse exit information
Get PDFEstimation of the optical properties of scattering media, such as tissue, is important in diagnostics as well as in the development of techniques to image deeper. Inverting scattering patterns to recover optical properties is not simple, and machine learning has been proposed to recover these properties. We train a neural network on simulated data to predict scattering (µs), reduced scattering (′), and absorption (µa) coefficients, as well as anisotropy factor ( g). Our best network achieves prediction Mean Absolute Relative Error (MARE) of 3.4% for µa and 2.1% for ′. We show that combining photon exit angle and position improves accuracy by ∼34% compared to using exit position alone. To capture angle information practically, we propose a novel, to our knowledge, method using intensity distributions measured at two planes above the sample. We also analyze the conditions when µs and g can be separated from ′. In the non-diffuse regime, g and µs are determined with MARE <10% in the majority of cases. In the diffusive regime, g and µs cannot be determined well. However, when tissue thickness is <2 transport mean free paths (TMFP), the majority of prediction MARE is <25%
TBB-Based AC / DC Modular Converter Design for High-Power Electrified Aircraft
Get PDFThis paper presents a Technical Building Block (TBB)-based converter architecture for a megawatt range power system with a high-voltage DC distribution in More Electric Aircraft (MEA). By arranging these units in different series and / or parallel configurations, the system achieves the required current, voltage, and power levels, ensuring a modular and scalable design. Various series and / or parallel module configurations have been investigated. Furthermore, various DC / AC converter topologies were evaluated by means of simulation to select the best candidate, focusing on power losses, cooling requirements, and power density. Based on the findings, the Active neutral point clamped (ANPC) converter is recommended, due to its better efficiency and power density. Additionally, experimental validation was conducted on the demonstrator converter modules. The influence of series and parallel connections on performance was also examined experimentally, supporting theoretical discussions
RISC Fault of Non-Salient Wound-Rotor Synchronous Generator: A Detection Method and Effect Analysis of Insulation
No full textRotor interturn short circuit (RISC) fault is one of the most prevalent fault types in the non-salient wound-rotor synchronous generator (WRSG). In the paper, the theoretical mechanism of the RISC fault is studied in detail. Building upon this understanding, from the fault detection and harm consequences, the further work of this article includes proposing a RISC detection method and analyzing the influence of RISC on the insulation thermal damage. On the one hand, a new detection for RISC fault is first proposed based on the phase voltage. The proposed detection method innovatively allows for accurate determination of both the severity and location of RISC faults. On the other hand, the loss and the temperature of the stator core-winding system are comprehensively investigated with varied RISC degrees and positions. The thermal responses of the winding insulation are obtained under the joint action of the external magnetic heat source and the internal electric heat source. The study identifies critical points of insulation vulnerability, particularly noting areas prone to noise and connection joints. This research contributes to the timely online identification of RISC faults and supports preventive maintenance strategies for protecting winding insulation. © 2024 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC
StockMARL: A Novel Multi-Agent Reinforcement Learning System to Dynamically Improve Trading Strategies
Get PDFContemporary financial markets, shaped by geopolitical tensions, technological innovation, and economic shocks, require adaptive trading strategies. We propose a novel framework that integrates multi-agent simulation with Reinforcement Learning (RL), allowing RL agents to acquire trading strategies by observing heterogeneous rule-based and reactive agents that emulate real-world investor behaviors such as day trading, momentum chasing, and risk aversion. To evaluate this approach, we introduce StockMARL, a multi-agent RL stock market simulation platform. Experimental results show that our observational diversity learning approach has advantages compared to models trained on historical price data and other MARL methods, with RL agents effectively deriving resilient strategies by learning from diverse reactive agent behaviors
Preferences and Listening Efficiency of Adults With Cochlear Implants During Online Communication
Get PDFObjectives:In recent years, there has been a profound increase in the use of remote online communication as a supplement to, and in many cases a replacement for, in-person interactions. While online communication tools hold potential to improve accessibility, previous studies have suggested that increased reliance on remote communication poses additional challenges for people with hearing loss, including those with a cochlear implant (CI). This study aimed to investigate the preferences and speech-reception performance of adults with a CI during online communication. Design: An online experiment was designed to study user preferences and listening efficiency during online communication under three presentation modes: "audio"(audio only), "video"(audio-visual), and "captions"(audio-visual plus captions). Fifty adults with at least 1 CI and 50 approximately age-matched people with acoustic hearing (AH) who did not use any hearing devices participated in the experiment. While viewing examples of prerecorded online conversations, participants were asked to indicate which presentation mode they would have preferred had they been participating in those online conversations themselves, and to explain why. Subsequently, participants performed a behavioral test of speech recognition using the same three presentation modes. A joint analysis of accuracy and response time was performed using a hierarchical linear ballistic accumulator model of choice decision-making to derive a measure of listening efficiency. Results: While people with AH preferred traditional audio-visual presentation, most people with a CI preferred the caption presentation mode, reporting that this helped them follow the conversations more easily. However, participants also reported some perceived drawbacks of captions, related to having to attend to an additional stream of information. In the speech-recognition test, people with AH achieved similar performance across all three presentation modes, while for people with a CI listening efficiency was enhanced by the addition of visual cues: relative to audio-only presentation, people with a CI benefited from the addition of video (i.e., being able to see the talker's face), and benefited further from the addition of captions. However, even with captions, listening efficiency for people with a CI remained below that achieved by people with AH. Conclusions:Captions to online conversations are valued by, and beneficial to, people with a CI, but are not without some drawbacks. Lower listening efficiency for people with a CI, even when captions are enabled, may be one reason behind reports in the literature that people with a CI often find online communication to be a cognitively effortful task. Further research is warranted into solutions to help close the performance gap that exists between people with a CI and their AH peers during online communication
Interpretative resources within interactions: the transformation of scientific research along a pathway of science communication
No full textMisinformation about science is becoming increasingly common both through passive transmission and through active weaponization by bad faith actors around the globe. Science communicators and educators should play a central role in combatting this but to do so they need a deeper understanding of how the meaning of scientific research is transformed as it is disseminated to the public. To help with this understanding, I present a narrative case study that follows a single piece of scientific research from original publication through interactions between the authoring scientist and a news reporter to its presentation in a popular news story read by a member of the public. By analyzing how fine-grained knowledge resources are activated in each step of this journey, this study demonstrates that differences between scientific and lay interpretation of scientific research is not the result of the public’s lack of knowledge but rather an emergent property of small shifts in meaning in each interaction and text. Furthermore, I argue that both scientists and non-scientists rely on fundamentally similar interpretative resources depending on the interactional context of use. This provides new analytical and theoretical directions of developing research to understand and tools to facilitate science communication and education
Embedding employability in the psychology undergraduate research project through a framework of authentic assessment and training support
Get PDFWe present an evaluation study of a framework of authentic assessment and training support in the undergraduate research project designed to provide students with opportunities to develop professional competencies through their research project, beyond core research-related skills, and with particular focus on reflection, to gain experience in identifying and articulating these competencies, and to support them in their project work. We present our approach to module design and key findings from the following measures: (a) A survey with final year students (N=201), (b) students’ self-evaluations of their proficiency in key competencies (N=251). Ourfindings show that engaging with the assessment supported students in developing key professional competencies, including reflection. Engaging with training supported students in assessment preparation, skills development, and self-efficacy. These findings provide evidence that designing the core curriculum with a focus on graduate competencies has a positive impact on student learning and skills development
Chromosome-Level Genome Assembly of the Asian Tramp Snail Bradybaena similaris (Stylommatophora: Camaenidae)
Get PDFWhile terrestrial land snails have long been used for evolutionary research, a lack of high-quality genomic resources has impeded recent progress. Bradybaena snails in particular have numerous intriguing traits that make them a good model for studying evolution, including shell pattern polymorphism and convergent evolution. They are also introduced and invasive across the world. In this study, we present a chromosome-level genome assembly of the Asian tramp snail Bradybaena similaris, utilizing 88-fold Illumina short-read sequences, 125-fold Nanopore long-read sequences, 63-fold PacBio HiFi sequences, and 47-fold Hi-C sequences. The assembled genome of 2.18 Gb is anchored to 28 chromosomes and exhibits high completeness (single copy, 91.7%; duplicates, 7.1%) and contiguity (N50 of 75.6 Mb). Additionally, we also obtained a high-quality transcriptome for annotation. This resource represents the first chromosome-level assembly for snails in the superfamily Helicoidea, which includes more than 5,000 species of terrestrial snails, and will facilitate genomic study in Bradybaena and, more broadly, in the superfamily Helicoidea
Social prescribing in primary care for people living with dementia: a qualitative exploration of different roles and services in England
Get PDFBackground: Dementia is a global public health challenge with the number of people living with the condition rapidly rising. Social prescribing in primary care has emerged as a person-centred approach connecting individuals with community support. It is increasingly explored for its potential to support people with complex needs, yet its role in dementia care remains uncertain. This study aimed to explore current provision of social prescribing for people living with dementia across England, identifying relevant aspects for dementia care, with particular focus on generic and specialist services. Methods: Semi-structured interviews were conducted with regional leads of social prescribing services and social prescribing link workers (SPLWs) across England. Data were analysed using template analysis to identify key themes. Results: Twenty-two participants were interviewed: ten social prescribing regional leads, four generic SPLWs commissioned to work with people aged > 18 years, and eight SPLWs working exclusively or partly with a specific adult population. Four themes were identified: family carer engagement key to supporting people living with dementia; service rather than person-centred care; the dominance of dementia in influencing support; and strategies for success: dementia centred social prescribing. Participants identified the central role of family carers in facilitating access to social prescribing, highlighting that carer support was often essential. Generic social prescribing frequently followed a ‘service-led’ approach, with service constraints negatively influencing interactions. Dementia was often perceived as the dominant support need, potentially marginalising individuals within broader social prescribing services. Despite these challenges, participants with more role flexibility, and/or more experience of dementia, demonstrated a range of successful strategies, illustrating the potential of social prescribing for people living with dementia. Conclusion: SPLWs perceive that social prescribing has potential to play a key role in support for people living with dementia and family carers. While its core principles align well with dementia care, our findings suggest a social prescribing model more tailored to the needs of people living with dementia, or additional dementia-specific training for generic SPLWs, may prove more effective. Further research is needed to assess the impact of these approaches, especially for people living with dementia without access to family carers
Designing open quantum systems for enabling quantum enhanced sensing through classical measurements
Get PDFQuantum systems in nonequilibrium conditions, where coherent many-body interactions compete with dissipative effects, can feature rich phase diagrams and emergent critical behavior. Associated collective effects, together with the continuous observation of quanta dissipated into the environment - typically photons - allow one to achieve quantum-enhanced parameter estimation. However, protocols for tapping this enhancement typically involve intricate measurements on the combined system-environment state. Here, we show that many-body quantum enhancement can in fact be obtained through classical measurements such as photon counting and homodyne detection. We illustrate this in detail for a class of open spin-boson models, which can be realized in trapped-ion or cavity QED setups. Our findings highlight a route toward the design of systems that enable a practical implementation of quantum-enhanced metrology through continuous classical measurements
Access Repository Dashboard
Do you manage Open Research Online? Become a CORE Member to access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard! 👇