26838 research outputs found
Sort by
Catalytic entanglement transformations with noisy hardware
Get PDFThe availability of certain entangled resource states (catalyst states) can enhance the rate of converting several less entangled states into fewer highly entangled states in a process known as catalytic entanglement concentration (EC). Here, we extend catalytic EC from pure states to mixed states and numerically benchmark it against non-catalytic EC and distillation in the presence of state-preparation errors and operational errors. Furthermore, we analyse the re-usability of catalysts in the presence of such errors. To do this, we introduce a novel recipe for determining the positive-operator valued measurements (POVM) required for EC transformations, which allows for making tradeoffs between the number of communication rounds and the number of auxiliary qubits required. We find that in the presence of low operational errors and depolarising noise, catalytic EC can provide better rates than distillation and non-catalytic EC
Generalised Invisible Markup
No full textInvisible XML makes the implicit structure of textual documents explicit by parsing, and then transforming the resultant parsetree into an abstract document. This abstract document is the essence of Invisible XML: it can be processed in many ways, though principally by serialising to XML.
However, as shown in an earlier paper on roundtripping ixml, it is possible to simultaneously simplify and generalise the ixml serialisation process, thereby opening it to other serialisations that are not hard-wired in the processor. By the same token, this further simplifies ixml proper, by reducing it to a simple transformation of grammars from ixml to an equivalent Invisible Markup grammar.
This talks discusses the changes needed to create a generalised Invisible Markup Language, explores the alternatives, and proposes future steps
Can you feel my brain? Investigating attentional engagement through EEG in an immersive musical environment
Get PDFIn de pers Robert Vis wil klanten CM afpakken & inflatie lijkt eindelijk iets te kantelen - Quotenet.nl - 2 December 2025
No full textFrom individual QoE to shared mental models: A novel evaluation paradigm for collaborative XR
Get PDFExtended Reality (XR) systems are rapidly shifting from isolated, single-user applications towards collaborative and social multi-user experiences. To evaluate the quality and effectiveness of such interactions, it is therefore required to move beyond traditional individual metrics such as Quality-of-Experience (QoE) or Sense of Presence (SoP). Instead, group-level dynamics such as effective communication, coordination etc. need to be encompassed to assess the shared understanding of goals and procedures. In psychology, this is referred to as a Shared Mental Model (SMM). The strength and congruence of such an SMM are known to be key for effective team collaboration and performance. In an immersive XR setting, though, novel Influence Factors (IFs) emerge that are not considered in a setting of physical co-location. Evaluations on the impact of these novel factors on SMM formation in XR, however, are close to non-existent. Therefore, this work proposes SMMs as a novel evaluation tool for collaborative and social XR experiences. To better understand how to explore this construct, we ran a prototypical experiment based on ITU recommendations in which the influence of asymmetric end-to-end latency is evaluated through a collaborative, two-user block building task. The results show how also in an XR context strong SMM formation can take place even when collaborators have fundamentally different responsibilities and behavior. Moreover, the study confirms previous findings by showing in an XR context that a teams’ SMM strength is positively associated with its performance
Conformalized polynomial chaos expansion for uncertainty-aware surrogate modeling
No full textThis work introduces a method to equip data-driven polynomial chaos expansion surrogate models with intervals that quantify the predictive uncertainty of the surrogate. To that end, jackknife-based conformal prediction is integrated into regression-based polynomial chaos expansions. The jackknife algorithm uses leave-one-out residuals to generate predictive intervals around the predictions of the polynomial chaos surrogate. The jackknife+ extension additionally requires leave-one-out model predictions. Both methods allow to use the entire dataset for model training and do not require a hold-out dataset for prediction interval calibration. The key to efficient implementation is to leverage the linearity of the polynomial chaos regression model, so that leave-one-out residuals and, if necessary, leave-one-out model predictions can be computed with analytical, closed-form expressions. This eliminates the need for repeated model re-training. The conformalized polynomial chaos expansion method is first validated on four benchmark models and then applied to two electrical engineering design use-cases. The method produces predictive intervals that provide the target coverages, even for low-accuracy models trained with small datasets. At the same time, training data availability plays a crucial role in improving the empirical coverage and narrowing the predictive interval, as well as in reducing their variability over different training datasets
Infill asymptotics for logistic regression estimators for parameters of the intensity function of spatial point processes
No full textThis paper discusses infill asymptotics for logistic regression estimators for spatial point processes whose intensity functions are of log-linear form. First, we establish strong consistency and asymptotic normality for the parameters of a Poisson point process model. We also propose consistent estimators for the asymptotic covariance matrix. Next, we extend the results to general point process models for which replicated realizations are available and, under proper conditions, extend the central limit theorem to estimators from other unbiased estimating equations that are based on the Campbell–Mecke theorem. In a simulation study, we demonstrate the efficiency of a regular dummy point process in logistic regression estimation and pseudo-likelihood estimation. Finally, we demonstrate the approach on data on kitchen fires in the Twente region in the Netherlands
Static X-ray CT of mixed-brand spaghetti bundles
No full textThe dataset consists of scans of 2 different pasta bundles of different brands placed inside a bottle. Each bundle is a mixture of specific brands where <em>Mix_All</em> consists of Albert Heijn - Spelt, Albert Heijn - Durum, DeCecco, and Gran Italia, while <em>Mix_DeCeccoSpelt</em> is a mixture of De Cecco and Albert Heijn - Spelt. The pasta was suspended at the bottleneck using a sponge to hold it in position. Scanning was focused on the bottleneck region. This scan was performed under static condition: the pasta is suspended in air.
The static scan of each of the spaghetti brand can be found here: https://zenodo.org/records/17312653">10.5281/zenodo.17312653
The dynamic counterpart of a spaghetti pasta can be found here: https://zenodo.org/records/15688351">10.5281/zenodo.15688351
Deep learning for next-generation communication technologies
Get PDFThis thesis presents a set of works on the use of deep learning (DL) in the next generation of communications systems. Digital communications is a cornerstone technology of the modern information age. Thus, a natural question to ask is whether DL enables better solutions for common problems in communications systems. To that end, this thesis poses two research questions: RQ1: How can deep learning techniques be effectively applied across different parts of the networking stack? RQ2: What are the advantages and challenges of implementing deep learning in the physical layer versus the application layer? This thesis consists of two parts with three chapters each. The first three content chapters discuss works that relate to the physical layer of the networking stack. The second three chapters contain works that relate to the application layer of the networking stack. Chapter 2 focuses on improving deep learning in the physical layer of communication networks. We develop a more flexible neural receiver that can handle multiple types of modulations. A modulation type is a configuration that makes a trade-off between speed and reliability. This approach makes deep learning based communication systems modular, making them more adaptable and efficient. Chapter 3 continues work on the physical layer, tackling the challenge of making deep learning receivers faster and more energy-efficient. We use a method of deep learning called group equivariant deep learning to build neural networks that inherently understand certain properties of radio signals. This results in smaller, more efficient networks that perform just as well as larger ones. Chapter 4 shifts to underwater acoustics. We use a technique called contrastive learning to train neural networks without labeled data. This approach could improve underwater sound classification and potentially help simulate underwater communication channels more accurately. Chapter 5 explores a security vulnerability in video streaming platforms like YouTube. Videos are generally streamed dynamically which is known to result in vulnerabilities. Using deep learning techniques, specifically deep-metric learning, we demonstrate that video IDs can be identified from encrypted streams with minimal examples. This work highlights how DL simplifies exploiting vulnerabilities between communication layers. Chapter 6 focuses on improving neural image compression. Building upon the work of Yang et al. (2020), we propose SGA+, a set of functions that improve the weighting of Gumbel probabilities. Our approach, particularly the SSL function with hyperparameter a, converges faster and achieves better PSNR/BPP trade-offs than the original SGA method. This advancement in neural image compression is important for reducing internet congestion and improving edge device capabilities. Chapter 7 investigates the application of Federated Learning (FL) in Face Recognition (FR) to address privacy concerns and issues with data heterogeneity. Given the challenges of non-shared identities across parties, we propose using federated meta-learning. Our approach demonstrates improved overall performance per client, especially under heterogeneous data splits. Notably, the performance gains primarily benefit weaker clients, reducing the variance in performance across clients. Together, these works provide partial answers to the research questions posed above. Major factors in applying DL across network layers are modularity, translating DL research to communications, and integrating DL advancements into the communications stack of tomorrow. Notably, there are differences between the application layer and physical layer. This thesis found the physical layer to benefit more broadly from the inclusion of DL. However, the models used throughout this work were highly similar; they were all based on convolutional neural networks (CNNs). The future of DL for communications looks bright and much work remains to be done