Last 7 days
Bookmark and Share    
Full Text
Peer Reviewed
See detailMO-Gym: A Library of Multi-Objective Reinforcement Learning Environments
Alegre, Lucas Nunes; Felten, Florian UL; Talbi, El-Ghazali UL et al

Scientific Conference (2022, November)

Detailed reference viewed: 31 (7 UL)
Full Text
Peer Reviewed
See detailDéploiement de l’espace muséal et engagement participatif des visiteurs : l’expérience perceptive d’une statue en ronde-bosse
Blanc, Mathias UL

in Les Cahiers de l'Ecole du Louvre (2022), 19

This article reports on fieldwork conducted at the Musée du Louvre-Lens with visitors discovering a sculpture from the Louvre, the Discophorus. To study the statue’s reception, we developed a mixed ... [more ▼]

This article reports on fieldwork conducted at the Musée du Louvre-Lens with visitors discovering a sculpture from the Louvre, the Discophorus. To study the statue’s reception, we developed a mixed methodology based on video recordings and “annotation traces in augmented reality”. The article presents the salient results of several weeks of fieldwork with and without the digital device Ikonikat 3D. It appears that our study corroborates the strong influence of the label on the way visitors look at the sculptures. But it also shows that the use of a digital device, depending on the human-machine interaction it supports, creates a space of vision and a space of action that have consequences in terms of distance and point of view on the work being viewed. Finally, we can observe different registers of meaning in the museum situation. This leads us to use a certain perspective to accompany visitor engagement with the sculptures. [less ▲]

Detailed reference viewed: 10 (0 UL)
Full Text
Peer Reviewed
See detailAdapting to Dynamic LEO-B5G Systems: Meta-Critic Learning Based Efficient Resource Scheduling
Yuan, Yaxiong; Lei, Lei; Vu, Thang Xuan UL et al

in IEEE Transactions on Wireless Communications (2022), 21(11), 9582-9595

Low earth orbit (LEO) satellite-assisted communications have been considered as one of the key elements in beyond 5G systems to provide wide coverage and cost-efficient data services. Such dynamic space ... [more ▼]

Low earth orbit (LEO) satellite-assisted communications have been considered as one of the key elements in beyond 5G systems to provide wide coverage and cost-efficient data services. Such dynamic space-terrestrial topologies impose an exponential increase in the degrees of freedom in network management. In this paper, we address two practical issues for an over-loaded LEO-terrestrial system. The first challenge is how to efficiently schedule resources to serve a massive number of connected users, such that more data and users can be delivered/served. The second challenge is how to make the algorithmic solution more resilient in adapting to dynamic wireless environments. We first propose an iterative suboptimal algorithm to provide an offline benchmark. To adapt to unforeseen variations, we propose an enhanced meta-critic learning algorithm (EMCL), where a hybrid neural network for parameterization and the Wolpertinger policy for action mapping are designed in EMCL. The results demonstrate EMCL’s effectiveness and fast-response capabilities in over-loaded systems and in adapting to dynamic environments compare to previous actor-critic and meta-learning methods. [less ▲]

Detailed reference viewed: 9 (0 UL)
Full Text
Peer Reviewed
See detailReSQ: Reinforcement Learning-Based Queue Allocation in Software-Defined Queuing Framework
Maity, Ilora UL; Taleb, Tarik

in Journal of Networking and Network Applications (2022), 2(4), 143152

With the evolution of 5G networks, the demand for Ultra-Reliable Low Latency Communications (URLLC) services is increasing. Software-Defined Networking (SDN) offers flexible network management to ... [more ▼]

With the evolution of 5G networks, the demand for Ultra-Reliable Low Latency Communications (URLLC) services is increasing. Software-Defined Networking (SDN) offers flexible network management to prioritize URLLC services coexisting with best-effort traffic. Utilizing the network programmability feature of SDN, Software-Defined Queueing (SDQ) framework selects the optimal output port queue on forwarding devices and routing path for incoming traffic flows to provide deterministic Quality of Service (QoS) support required for URLLC traffic. However, in the existing SDQ framework, the selections of optimal queue and path are done manually by observing the traffic type of each incoming flow, the available bandwidth of each potential routing path, and the status of output port queues of each forwarding device on each potential routing path. The static allocations of path and queue for each flow are inefficient to provide a deterministic QoS guarantee for a high volume of incoming traffic which is typical in 5G networks. The limited buffer availability on the forwarding devices is another constraint regarding optimal queue allocation that ensures an end-to-end (E2E) delay guarantee. To address these challenges, in this paper, we extend the SDQ framework by automating queue management with a reinforcement learning (RL)-based approach. The proposed queue management approach considers diverse QoS demands as well as a limited buffer on the forwarding devices and performs prioritized queue allocation. Our approach also includes a hash-based flow grouping to handle a high volume of traffic having diverse latency demands and a path selection mechanism based on available bandwidth and hop count. The simulation result shows that the proposed scheme ReSQ reduces the QoS violation ratio by 10.45% as compared to the baseline scheme that selects queues randomly. [less ▲]

Detailed reference viewed: 17 (6 UL)
See detailJack of All Trades - Stolz und Hilflosigkeit
Richter, Daniel UL

Article for general public (2022)

Detailed reference viewed: 40 (1 UL)
Full Text
Peer Reviewed
See detailCalcGraph: taming the high costs of deep learning using models
Lorentz, Joe UL; Hartmann, Thomas UL; Moawad, Assaad UL et al

in Software and Systems Modeling (2022)

Models based on differential programming, like deep neural networks, are well established in research and able to outperform manually coded counterparts in many applications. Today, there is a rising ... [more ▼]

Models based on differential programming, like deep neural networks, are well established in research and able to outperform manually coded counterparts in many applications. Today, there is a rising interest to introduce this flexible modeling to solve real-world problems. A major challenge when moving from research to application is the strict constraints on computational resources (memory and time). It is difficult to determine and contain the resource requirements of differential models, especially during the early training and hyperparameter exploration stages. In this article, we address this challenge by introducing CalcGraph, a model abstraction of differentiable programming layers. CalcGraph allows to model the computational resources that should be used and then CalcGraph’s model interpreter can automatically schedule the execution respecting the specifications made. We propose a novel way to efficiently switch models from storage to preallocated memory zones and vice versa to maximize the number of model executions given the available resources. We demonstrate the efficiency of our approach by showing that it consumes less resources than state-of-the-art frameworks like TensorFlow and PyTorch for single-model and multi-model execution. [less ▲]

Detailed reference viewed: 23 (1 UL)
Full Text
See detailForecasting of a complex microbial community using meta-omics
Delogu, Francesco UL; Kunath, Benoît UL; Queirós, P. M. et al

E-print/Working paper (2022)

Microbial communities are complex assemblages whose dynamics are shaped by abiotic and biotic factors. A major challenge concerns correctly forecasting the community behaviour in the future. In this ... [more ▼]

Microbial communities are complex assemblages whose dynamics are shaped by abiotic and biotic factors. A major challenge concerns correctly forecasting the community behaviour in the future. In this context, communities in biological wastewater treatment plants (BWWTPs) represent excellent model systems, because forecasting them is required to ultimately control and operate the plants in a sustainable manner. Here, we forecast the microbial community from the water-air interface of the anaerobic tank of a BWWTP via longitudinal meta-omics (metagenomics, metatranscriptomics and metaproteomics) data covering 14 months at weekly intervals. We extracted all the available time-dependent information, summarised it in 17 temporal signals (explaining 91.1 of the temporal variance) and linked them over time to rebuild the sequence of ecological phenomena behind the community dynamics. We forecasted the signals over the following five years and tested the predictions with 21 extra samples. We were able to correctly forecast five signals accounting for 22.5 of the time-dependent information in the system and generate mechanistic predictions on the ecological events in the community (e.g. a predation cycle involving bacteria, viruses and amoebas). Through the forecasting of the 17 signals and the environmental variables readings we reconstructed the gene abundance and expression for the following 5 years, showing a nearly perfect trend prediction (coefficient of determination >= 0.97) for the first 2 years. The study demonstrates the maturity of microbial ecology to forecast composition and gene expression of open microbial ecosystems using year-spanning interactions between community cycles and environmental parameters. [less ▲]

Detailed reference viewed: 41 (4 UL)
Full Text
See detailWohlbefinden in der Schule: Aktuelle Befunde, Potenziale und Herausforderungen
Residori, Caroline UL; Heinen, Andreas UL; Samuel, Robin UL

Conference given outside the academic context (2022)

Detailed reference viewed: 12 (1 UL)
See detailEnvelhecimento Saudável O que fazer para buscarmos envelhecer com saúde ?
Teixeira Santos, Ana Carolina UL

Conference given outside the academic context (2022)

Detailed reference viewed: 22 (0 UL)
Full Text
Peer Reviewed
See detailDBSegment: Fast and robust segmentation of deep brain structures considering domain generalisation
Baniasadi, Mehri UL; Petersen, Mikkel V.; Goncalves, Jorge UL et al

in Human Brain Mapping (2022)

Segmenting deep brain structures from magnetic resonance images is important for patient diagnosis, surgical planning, and research. Most current state-of-the-art solutions follow a segmentation-by ... [more ▼]

Segmenting deep brain structures from magnetic resonance images is important for patient diagnosis, surgical planning, and research. Most current state-of-the-art solutions follow a segmentation-by-registration approach, where subject magnetic resonance imaging (MRIs) are mapped to a template with well-defined segmentations. However, registration-based pipelines are time-consuming, thus, limiting their clinical use. This paper uses deep learning to provide a one-step, robust, and efficient deep brain segmentation solution directly in the native space. The method consists of a preprocessing step to conform all MRI images to the same orientation, followed by a convolutional neural network using the nnU-Net framework. We use a total of 14 datasets from both research and clinical collections. Of these, seven were used for training and validation and seven were retained for testing. We trained the network to segment 30 deep brain structures, as well as a brain mask, using labels generated from a registration-based approach. We evaluated the generalizability of the network by performing a leave-one-dataset-out cross-validation, and independent testing on unseen datasets. Furthermore, we assessed cross-domain transportability by evaluating the results separately on different domains. We achieved an average dice score similarity of 0.89 ± 0.04 on the test datasets when compared to the registration-based gold standard. On our test system, the computation time decreased from 43 min for a reference registration-based pipeline to 1.3 min. Our proposed method is fast, robust, and generalizes with high reliability. It can be extended to the segmentation of other brain structures. It is publicly available on GitHub, and as a pip package for convenient usage. [less ▲]

Detailed reference viewed: 10 (0 UL)
Full Text
Peer Reviewed
See detailAutomatic Repair and Deadlock Detection for Parameterized Systems
Jacobs, Swen; Sakr, Mouhammad UL; Volp, Marcus UL

in Automatic Repair and Deadlock Detection for Parameterized Systems (2022, October 15)

Detailed reference viewed: 11 (0 UL)
Full Text
See detailInduced superconductivity in a quantum Hall edge state
Michelsen, Andreas Nicolai Bock UL

Doctoral thesis (2022)

In the search for non-Abelian anyonic zero modes for inherently fault-tolerant quantum computing, the hybridized superconductor - quantum Hall edge system plays an important role. Inspired by recent ... [more ▼]

In the search for non-Abelian anyonic zero modes for inherently fault-tolerant quantum computing, the hybridized superconductor - quantum Hall edge system plays an important role. Inspired by recent experimental realizations of this system, we describe it through a microscopic theory based on a BCS superconductor with Rashba spin-orbit coupling and Meissner effect at the surface, which is tunnel-coupled to a spin-polarized integer or fractional quantum Hall edge. By integrating out the superconductor, we arrive at an effective theory of the proximitized edge state and establish a qualitative description of the induced superconductivity. We predict analytical relations between experimentally available parameters and the key parameters of the induced superconductivity, as well as the experimentally relevant transport signatures. Extending the model to the fractional quantum Hall case, we find that both the spin-orbit coupling and the Meissner effect play central roles. The former allows for transport across the interface, while the latter controls the topological phase transition of the induced p-wave pairing in the edge state, allows for particle-hole conversion in transport for weak induced pairing amplitudes, and determines when pairing dominates over fractionalization in the proximitized fractional quantum Hall edge. Further experimental indicators are predicted for the system of a superonductor coupled through a quantum point contact with an integer or fractional quantum Hall edge, with a Pauli blockade which is robust to interactions and fractionalization as a key indicator of induced superconductivity. With these predictions we establish a more solid qualitative understanding of this important system, and advance the field towards the realization of anyonic zero modes. [less ▲]

Detailed reference viewed: 9 (1 UL)
Full Text
Peer Reviewed
See detailAn Evolutionary Algorithm to Optimise a Distributed UAV Swarm Formation System
Stolfi Rosso, Daniel UL; Danoy, Grégoire UL

in Applied Sciences (2022), 12(20),

In this article, we present a distributed robot 3D formation system optimally parameterised by a hybrid evolutionary algorithm (EA) in order to improve its efficiency and robustness. To achieve that, we ... [more ▼]

In this article, we present a distributed robot 3D formation system optimally parameterised by a hybrid evolutionary algorithm (EA) in order to improve its efficiency and robustness. To achieve that, we first describe the novel distributed formation algorithm3 (DFA3), the proposed EA, and the two crossover operators to be tested. The EA hyperparameterisation is performed by using the irace package and the evaluation of the three case studies featuring three, five, and ten unmanned aerial vehicles (UAVs) is performed through realistic simulations by using ARGoS and ten scenarios evaluated in parallel to improve the robustness of the configurations calculated. The optimisation results, reported with statistical significance, and the validation performed on 270 unseen scenarios show that the use of a metaheuristic is imperative for such a complex problem despite some overfitting observed under certain circumstances. All in all, the UAV swarm self-organised itself to achieve stable formations in 95% of the scenarios studied with a plus/minus ten percent tolerance. [less ▲]

Detailed reference viewed: 8 (0 UL)
Full Text
Peer Reviewed
See detailEugenia, incarnée par Marina Hands, dans Actrice de Pascal Rambert : le corps d’une mourante comme figuration de la Dame Blanche
Deregnoncourt, Marine UL

in Laforêt, Carole; Raimbault, Nicole (Eds.) Les femmes et leur corps (2022, October 10)

Soit cette description physique de l’actrice franco-britannique Marina Hands par Éric Ruf (Administrateur Général de la Comédie-Française depuis 2014) : Marina est une jeune femme très belle, mais elle a ... [more ▼]

Soit cette description physique de l’actrice franco-britannique Marina Hands par Éric Ruf (Administrateur Général de la Comédie-Française depuis 2014) : Marina est une jeune femme très belle, mais elle a une beauté très singulière. Elle ne correspond en rien au canon. Elle a des épaules des nageuses est-allemandes, elle a des bras longs comme des kilomètres, elle est massive, elle a une drôle de tronche avec des pommettes extrêmement saillantes. Elle n’a pas une beauté classique. Nous avons conclu ainsi notre exposé au 10ème colloque orléanais consacré aux sorcières. En vue de contribuer humblement et d’apporter modestement notre pierre à l’édifice du prochain congrès intitulé : « Femmes des lumières et de l'ombre. Les Femmes et leur Corps », nous entendons, par le biais de cette nouvelle communication, prolonger notre réflexion initiale sur Marina Hands en nous axant plus particulièrement, cette fois-ci, sur le corps de cette comédienne au miroir de son interprétation d’Eugenia dans Actrice de Pascal Rambert. Tout d’abord, en quoi le corps de Marina Hands peut-il être défini comme « masculin » et quels en sont, non seulement, les enjeux, mais aussi, les implications sur son jeu d’actrice ? C’est précisément ce que nous entendons aborder, par la suite, avec le rôle d’Eugenia, protagoniste autour de laquelle se construit l’action d’Actrice de Pascal Rambert. Pourquoi le corps d’Eugenia, mourante, peut-il être perçu comme une figuration de la Dame Blanche, messagère de la mort, et comment Pascal Rambert, renforce-t-il cette idée en construisant un requiem, par un dispositif scénographique singulier ? C’est spécifiquement à ces trois questions centrales auxquelles cette présentation souhaite répondre [less ▲]

Detailed reference viewed: 24 (1 UL)
Full Text
Peer Reviewed
See detailOn the Performance of Cache-Free/Cache-Aided STBC-NOMA in Cognitive Hybrid Satellite-Terrestrial Networks
Singh, Vibhum UL; Solanki, Sourabh UL; Eappen, Geoffrey UL et al

in IEEE Wireless Communications Letters (2022), (Early Access), 1-1

Future wireless networks pose several challenges such as high spectral efficiency, wide coverage massive connectivity, low receiver complexity, etc. To this end, this letter investigates an overlay based ... [more ▼]

Future wireless networks pose several challenges such as high spectral efficiency, wide coverage massive connectivity, low receiver complexity, etc. To this end, this letter investigates an overlay based cognitive hybrid satellite-terrestrial network (CHSTN) combining non-orthogonal multiple access (NOMA) and conventional Alamouti space-time block coding (STBC) techniques. Herein, a decode-and-forward based secondary terrestrial network cooperates with a primary satellite network for dynamic spectrum access. Further, for reliable content delivery and low latency requirements, wireless caching is employed, whereby the secondary network can store the most popular contents of the primary network. Considering the relevant heterogeneous fading channel models and the NOMA-based imperfect successive interference cancellation, we examine the performance of CHSTN for the cache-free (CF) STBC-NOMA and the cache-aided (CA) STBC-NOMA schemes. We assess the outage probability expressions for primary and secondary networks and further, highlight the corresponding achievable diversity orders. Indicatively, the proposed CF/CA STBC-NOMA schemes for CHSTN perform significantly better than the benchmark standalone NOMA and OMA schemes. [less ▲]

Detailed reference viewed: 32 (17 UL)
Full Text
See detail3D Modelling for AR and 3D printing in Teacher Training
Ulbrich, Eva; El Bedewy, Shereen; Handl, Julia et al

Scientific Conference (2022, October 07)

School activities integrating students’ environments into teaching aim to develop skills and strategies to solve problems in real-world situations and can be useful in hybrid teaching. Such activities can ... [more ▼]

School activities integrating students’ environments into teaching aim to develop skills and strategies to solve problems in real-world situations and can be useful in hybrid teaching. Such activities can encourage and motivate exploring skills in Science, Technology, Engineering, Arts, and Mathematics (STEAM). Hybrid teaching usually uses technologies and connects virtual and physical worlds. We use technologies like 3D modelling for Augmented Reality (AR) or 3D printing with GeoGebra and created an exercise introducing them in a lecture for pre-service mathematics students. The exercise combines the possibility to introduce these technologies, can be used in hybrid teaching and connects to the Austrian mathematics curriculum. The exercise consists of 3D modelling mathematical mazes that can be explored using AR on handheld devices and can also be 3D printed. We used it in online, offline and hybrid scenarios with pre- and in-service teachers and will show resulting presentations of teacher projects. [less ▲]

Detailed reference viewed: 22 (5 UL)
Full Text
See detailAugmented Reality in primary education: New learning opportunities for students with learning difficulties in mathematics education
Haas, Ben; Kreis, Yves UL; Lavizcza, Zsolt

Scientific Conference (2022, October 06)

Although there are manifold connections between mathematics, foremost geometry, and the real-world (e.g., architecture, arts, functional objects), integration seldom happens in daily learning lessons in ... [more ▼]

Although there are manifold connections between mathematics, foremost geometry, and the real-world (e.g., architecture, arts, functional objects), integration seldom happens in daily learning lessons in mathematics primary education. Learning three-dimensional geometric shapes, for example, is mainly done in a two-dimensional setting using textbooks instead of three-dimensional settings using technology or didactical material. This circumstance, however, makes it far more difficult for students with learning difficulties in mathematics to understand mathematical properties, recognize shapes in the real world, and understand the possibilities of modulating shapes. Students with learning difficulties learn efficient strategies to apply mathematics to their environment when shapes and connections are visualized with Augmented Reality within the real world. Based on several experiences and studies, we will present and discuss learning mathematics with Augmented Reality in primary education for students with learning difficulties. [less ▲]

Detailed reference viewed: 16 (3 UL)
Full Text
Peer Reviewed
See detailTask-Oriented Data Compression for Multi-Agent Communications Over Bit-Budgeted Channels
Mostaani, Arsham UL; Vu, Thang Xuan UL; Chatzinotas, Symeon UL et al

in IEEE Open Journal of the Communications Society (2022)

Various applications for inter-machine communications are on the rise. Whether it is for autonomous driving vehicles or the internet of everything, machines are more connected than ever to improve their ... [more ▼]

Various applications for inter-machine communications are on the rise. Whether it is for autonomous driving vehicles or the internet of everything, machines are more connected than ever to improve their performance in fulfilling a given task. While in traditional communications the goal has often been to reconstruct the underlying message, under the emerging task-oriented paradigm, the goal of communication is to enable the receiving end to make more informed decisions or more precise estimates/computations. Motivated by these recent developments, in this paper, we perform an indirect design of the communications in a multi-agent system (MAS) in which agents cooperate to maximize the averaged sum of discounted one-stage rewards of a collaborative task. Due to the bit-budgeted communications between the agents, each agent should efficiently represent its local observation and communicate an abstracted version of the observations to improve the collaborative task performance. We first show that this problem can be approximated as a form of data-quantization problem which we call task-oriented data compression (TODC). We then introduce the state-aggregation for information compression algorithm (SAIC) to solve the formulated TODC problem. It is shown that SAIC is able to achieve near-optimal performance in terms of the achieved sum of discounted rewards. The proposed algorithm is applied to a geometric consensus problem and its performance is compared with several benchmarks. Numerical experiments confirm the promise of this indirect design approach for task-oriented multi-agent communications. [less ▲]

Detailed reference viewed: 30 (3 UL)
Full Text
Peer Reviewed
See detailCambridge hybrid closed-loop system in very young children with type 1 diabetes reduces caregivers’ fear of hypoglycemia and improves their well-being
de Beaufort, Carine UL; Schierloh, Ulrike; Thankamony, A. et al

in Diabetes Care (2022), 45

Objective To evaluate the impact of CamAPS FX hybrid closed-loop automated insulin delivery (HCL) in very young children with type 1 diabetes (T1D) on caregivers’ well-being, fear of hypoglycemia and ... [more ▼]

Objective To evaluate the impact of CamAPS FX hybrid closed-loop automated insulin delivery (HCL) in very young children with type 1 diabetes (T1D) on caregivers’ well-being, fear of hypoglycemia and sleepiness. Research Design Multinational, open label, randomized crossover study. Children (1-7years) with T1D received treatment for two 4-month periods in random order, comparing HCL with sensor augmented pump (SAP) (control). At baseline and after each treatment period, caregivers were invited to complete WHO-5, Hypoglycemia Fear Survey (HFS) and Epworth Sleepiness Scale (ESS). Results Caregivers of 74 children (mean±SD: age 5±2 years; 42% female, baseline HbA1c 7.3±0.7%) participated. Results revealed significantly lower scores for hypoglycemia fear (p<.001) and higher for well-being (p<.001) after HCL treatment. A trend towards a reduction in sleepiness score was observed (p=0.09). Conclusion Our results suggest a better well-being and less hypoglycemia fear in caregivers of very young children with T1D on CamAPS FX HCL. [less ▲]

Detailed reference viewed: 7 (0 UL)