Additive Schwarz methods for serendipity elements

Serendipity elements allow us to obtain the same order of accuracy as rectangular tensor-product elements with many fewer degrees offreedom (DOFs) To realize the possible computational savings, we develop someadditive Schwarz methods (ASM) based on solving local patch problems . We prove that patchsmoothers give conditioning estimates independent of the polynomial degree for a model problem .…

Robust Connectivity of Graphs on Surfaces

Large robust connectivity was originally used to show flexible choosability in non-regular graphs . In this paper, we investigate some interesting properties of robustconnectivity for graphs embedded in surfaces . We prove a tight asymptotic bound of $Omega for the robust connectivity of $r$-connectedgraphs of Euler genus $gamma .…

MultiCruise Eco Lane Selection Strategy with Eco Cruise Control for Connected and Automated Vehicles

Connected and Automated Vehicles (CAVs) have real-time information from the surrounding environment by using local on-board sensors, V2X(Vehicle-to-Everything) communications, pre-loaded vehicle-specific lookuptables, and map database . Eco-Cruise and Eco-LaneSelection on highways and/or motorways have immense potential to save energy,because there are generally fewer traffic controllers and the vehicles keepmoving in general .…

COMTEST Project A Complete Modular Test Stand for Human and Humanoid Posture Control and Balance

This work presents a system to benchmark humanoid posture control and balanceperformances under perturbed conditions . System includes a motion platform used toprovide the perturbation, an innovative body-tracking system suitable for robots, humans and exoskeletons . The design of the system is modularity: all its components can be replaced or extended according to experimental needs .…

MILIOM Tightly Coupled Multi Input Lidar Inertia Odometry and Mapping

In this paper we investigate a tightly coupled Lidar-Inertia Odometry andMapping (LIOM) scheme . We devise a time-synchronizedscheme to combine extracted features from separate lidars into a singlepointcloud . These coefficients, along with IMUpreinteration observations, are then used to construct a factor graph that will be optimized to produce an estimate of the sliding window trajectory .…

Constantine Automatic Side Channel Resistance Using Efficient Control and Data Flow Linearization

In the era of microarchitectural side channels, vendors scramble to deploymitigations for transient execution attacks . But traditional side-channel attacks against sensitive software (e.g., crypto programs) must be fixed by means of constant-time programming . Constantine pursues a radical design point where secret-dependent control and data flows are completely linearized (i.e.,…

EXplainable Neural Symbolic Learning X NeSyL methodology to fuse deep learning representations with expert knowledge graphs the MonuMAI cultural heritage use case

Deep Learning (DL) models for detection and classification have achieved an unprecedented performance over classical machine learning algorithms . However, DL models are black-box methods hard to debug, interpret, and certify . In contrast, symbolic AI systems that convert concepts into rules or symbols are easier to explain .…

Adaptive Sampling Algorithmic vs Human Waypoint Selection

A paper compares the performance of humans versus adaptiveinformative sampling algorithms for selecting informative waypoints . The results show that the robot can on average perform better than the average human, and approximately as good as the best human, when the model assumptions don’t correspond to the characteristics of the field .…

Explainable Artificial Intelligence Reveals Novel Insight into Tumor Microenvironment Conditions Linked with Better Prognosis in Patients with Breast Cancer

We investigated the relationship between features in the tumormicroenvironment (TME) and the overall and 5-year survival in triple-negativebreast cancer (TNBC) and non-TNBC (NTNBC) patients by using ExplainableArtificial Intelligence (XAI) models . Novel insights derived from our XAI model showed that CD4+ T cells and B cells are more critical than other TME features for enhanced prognosis for both TNBC and NTNBC patients .…

Anomaly Detection for Solder Joints Using β VAE

In the assembly process of printed circuit boards, most of the errors are caused by solder joints in Surface Mount Devices . Traditional feature extraction based methods require designing hand-craftedfeatures and rely on the tiered RGB illumination to detect solder joint errors .…

Ask Explore Grounded Question Answering for Curiosity Driven Exploration

In many real-world scenarios where extrinsic rewards to the agent are extremely sparse, curiosity has emerged as a useful concept . We show that natural language questions encourage the agent to uncoverspecific knowledge about their environment such as the physical properties of objects as well as their spatial relationships with other objects, which serve as valuable curiosity rewards to solve sparse-reward tasks more efficiently .…

Drone Launched Short Range Rockets

A concept of drone launched short range rockets (DLSRR) is presented . A drone or an aircraft rises DLSRR to a release altitude of up to 20 $km$. At therelease altitude, the drone or aircraft is moving at a velocity of 700$m/s$ and a steep angle of 68$^o$ to the horizontal .…

CFNet LiDAR Camera Registration Using Calibration Flow Network

LiDAR-camera calibration is critical for autonomous vehicles and robot navigation . EPnP algorithm within the RANdom SAmpleConsensus (RANSAC) scheme is applied to estimate the extrinsic parameters with2D-3D correspondences constructed by the calibration flow . We propose a semantic initializational algorithm with the introduction of instance centroids (ICs) The code will bepublicly available at https://://://github.com/LvXudong-HIT/CFNet.…

Aligned Contrastive Predictive Coding

Aligned Contrastive PredictiveCoding (ACPC) leads to higher linear phone prediction accuracy and lower ABXerror rates, while being slightly faster to train due to the reduced number of prediction heads . We evaluate the model on a speechcoding task and demonstrate that the proposed ACPC leads to .…

Wireless Federated Learning WFL for 6G Networks Part II The Compute then Transmit NOMA Paradigm

The Compute-then-Transmit NOMA(CT-NOMA) protocol is introduced, where users terminate concurrently the localmodel training and then simultaneously transmit the trained parameters to the central server . The simulation resultsverify the effectiveness of CT-NomA in terms of delay reduction, compared to the considered benchmark that is based on time-division multiple access .…

Membrane Fusion Based Transmitter Design for Static and Diffusive Mobile Molecular Communication Systems

This paper proposes a novel imperfect transmitter (TX) model that adopts MF between a vesicle and the TXmembrane to release molecules encapsulated within the vesicles . Incorporating molecular degradation and afully-absorbing receiver (RX), the channel impulse response (CIR) is derived for two scenarios: Both TX and RX are static, and both TX and TX arediffusion-based mobile .…

Suboptimal coverings for continuous spaces of control tasks

We propose the suboptimal covering number to characterize multi-task control problems where the set of dynamical systems and/or cost functions is infinite . This notion may help quantify the function class expressiveness needed to represent a goodmulti-task policy, which is important for learning-based control methods that use parameterized function approximation .…

Towards Low burden Responses to Open Questions in VR

Subjective self-reports in VR user studies is a burdening and often tedioustask for the participants . To minimize the disruption with the ongoingexperience VR research has started to administer the surveying directly insidethe virtual environments . However, due to the tedious nature of text-entry inVR, most VR surveying tools focus on closed questions with predetermined answers, while open questions with free-text responses remain unexplored .…

Social Influence Prediction with Train and Test Time Augmentation for Graph Neural Networks

Data augmentation has been widely used in machine learning for naturallanguage processing and computer vision tasks to improve model performance . But little research has studied data augmentation on graph neuralnetworks, particularly using augmentation at both train- and test-time . We have designed amethod for social influence prediction using graph neural networks with train-and-test-time augmentation, which can effectively generate multiple augmentedgraphs for social networks by utilising a variational graph autoencoder in both scenarios .…

SnapCheck Automated Testing for Snap Programs

Programming environments such as Snap, Scratch, and Processing engage students by allowing them to create programming artifacts such as apps and games, with visual and interactive output . Learning programming with such amedia-focused context has shown to increase retention and success rate .…

Intensional Artificial Intelligence From Symbol Emergence to Explainable and Empathetic AI

We argue that an explainable artificial intelligence must possess a rationalefor its decisions, be able to infer the purpose of observed behaviour . To communicate that rationale requires naturallanguage, a means of encoding and decoding perceptual states . We propose atheory of meaning in which, to acquire language, an agent should model the world a language describes rather than the language itself.…

Uniformly accurate low regularity integrators for the Klein Gordon equation from the classical to non relativistic limit regime

We propose a novel class of uniformly accurate integrators for the Klein–Gordon equation . They capture classical $c=1$ as well as highly-oscillatory non-relativistic regimes $c\gg1$ and, at the same time,allow for low regularity approximations . The schemes converge under lower regularityassumptions than classical schemes, such as splitting or exponential integratormethods, require .…

Convexification based globally convergent numerical method for a 1D coefficient inverse problem with experimental data

To solve the inverse problem, we establish a new version of Carlemanestimate and then employ this estimate to construct a cost functional which isstrictly convex on a convex bounded set with an arbitrary diameter in a Hilbertspace . Minimizing this convex functional by the gradient descent method, we obtain the desired numerical solution to the coefficient inverse problems .…

Reduced order models for Lagrangian hydrodynamics

Lagrangian hydrodynamics is characterized by moving meshes, advection-dominated solutions, and moving shock fronts withsharp gradients . These challenges hinder the existing projection-based modelreduction schemes from being practical . Over-sampling DEIM and SNS approaches reduce complexity due to the nonlinear terms .…

Discrete Maximum principle of a high order finite difference scheme for a generalized Allen Cahn equation

We consider solving a generalized Allen-Cahn equation coupled with a passive convection for a given incompressible velocity field . We prove that the discrete maximum principle holds under suitable meshsize and time step constraints . The same result also applies to construct abound-preserving scheme for any convection with an incompressablevelocity field, we say .…

Optimizing small BERTs trained for German NER

Currently, the most widespread neural network architecture for traininglanguage models is the so called BERT . In general, the larger the number of parameters in a BERT model, the results obtained in these NLP tasks . Unfortunately, the memory and training duration drastically increases with the size of these models .…

Understanding who uses Reddit Profiling individuals with a self reported bipolar disorder diagnosis

This paper shows how existing NLP methods can yield information on clinical, demographic, and identity characteristics of almost 20K Reddit users who self-report a bipolar disorder diagnosis . This population consists ofslightly more feminine- than masculine-gendered mainly young or middle-aged US-based adults who often report additional mental health diagnoses, which iscompared with general Reddit statistics and epidemiological studies .…

Claim Detection in Biomedical Twitter Posts

Social media contains unfiltered and unique information, which is potentially of great value, but can also do great harm . Methods of automatic fact-checking and fake news detection addressthere problem, but have not been applied to the biomedical domain in socialmedia yet .…

Weakly supervised Multi task Learning for Multimodal Affect Recognition

Multimodal affect recognition constitutes an important aspect for enhancinginterpersonal relationships in human-computer interaction . However, relevant data is hard to come by and notably costly to annotate, which poses a challenge to build robust systems . We propose to leverage these datasets using weakly-supervised multi-task learning to improve the generalization performance on each of them .…

Knodle Modular Weakly Supervised Learning with PyTorch

Methods for improving the training and prediction quality of weakly supervised machine learning models vary in how much they are tailored to aspecific task, or integrated with a specific model architecture . In this work, we propose a software framework Knodle that provides a modularization forseparating weak data annotations and powerful deep learning models .…