Transfer Learning for Human & AI

Evidential occupancy grid maps are a popular representation of theenvironment of automated vehicles . Inverse sensor models (ISMs) are used to compute OGMs from sensor data such as lidar point clouds . Geometric ISMs show alimited performance when estimating states in unobserved but inferable areas .…

Developers use and contribute to repositories on GitHub . Documentation present in the repositories serves as an important source by helping developersto understand, maintain and contribute . Currently, documentationin a repository is diversified, among various files, with most of it present inReadMe files .…

Control of the batch transesterification process is difficult due to its complex and non-lineardynamics . This study explores the feasibility of reinforcement learning (RL) based control of the process . The results showcase that TD3 based controller is able to control batchtransesterification process and can be a promising direction towards the goal of artificial intelligence-based control in process industries .…

A recently developed algorithm allows for optimal planning over long and infinite horizons for multiple independent tasks that are partially observable and evolve over time . We present the theoretical analysis and proofs of the algorithm .…

nsan uses compile-time instrumentation to augment each floating-pointcomputation in the program with a higher-precision shadow which is checked for consistency during program execution . This makes nsan between 1 and 4 orders ofmagnitude faster than existing approaches, which allows running it routinely as part of unit tests, or detecting issues in large production applications .…

The error representation using the straight difference of two vectors in theinertial navigation system may not be reasonable as it does not take the direction difference into consideration . We proposed to use the$SE_2(3)$ matrix Lie group to represent the state of the inertial-integrated navigation system .…

An aerial vehicle powered by flapping feathered wings was designed, developed and fabricated . Different from legacy flapping-wing aerial vehicles withmembrane wings, the new design uses authentic bird feathers to fabricate wings . In field tests, a radio-controlled electric-powered aerial vehicle successfully took off, flew up to 63.88 s and landedsafely .…

A cloud-based internetof things (CIoT) service is employed to connect, supervise and control a robot arm’s motion using the added wireless sensing devices to the environment . A programmable automation controller (PAC) unit, connected to the robot arm receives the most recent changes and updates the motion of the robotarm .…

Disjunctive hybrid MKNF knowledge bases succinctly extend ASP and insome cases without increasing the complexity of reasoning tasks . The only known solver development is based onguess-and-verify, as formulated by Motik and Rosati in their original work . We formalize a notion of unfoundedsets for these knowledge bases, identify lower complexity bounds, and demonstrate how we might integrate these developments into a solver .…

The problem involves planning motions for a pair of planar robots connected to oneanother via a cable of limited length . The paper shows how the reduced visibilitygraph provides a natural discretization and captures the essential topologicalconsiderations very effectively for the two robot case as well .…

Stochastic models capture uncertainty and risk tolerance that are often present in real systems of interest . This paper extends maximum likelihood constraint inference to stochastic applications by using maximum causal entropy likelihoods . We propose an efficient algorithm that computes constraint likelihood and risktolerance in a unified Bellman backup, allowing us to generalize to Stochasticsystems without increasing computational complexity .…

This paper presents a constraint management strategy based on ScalarReference Governors (SRG) to enforce output, state, and control constraints . The strategy, referred to as the Preview Reference Governor (PRG), can outperformSRG while maintaining the computational benefits of SRG . However, performance of PRG may suffer if large preview horizons are used .…

We propose a new method of generating meaningful embeddings for speech, changes to four commonly used meta learning approaches and an approach of combining their outcomes into an end-to-end automatic speech recognition system . We verify the functionality of each of our three contributions in two experiments exploring their performance for differentamounts of classes (N-way) and examples per class (k-shot) in a few-shotsetting .…

We derive novel results on the ergodic theory of irreducible, aperiodicMarkov chains . We show how to optimally steer the network flow to a stationary distribution over a finite or infinite time horizon . Optimality is with respectto an entropic distance between distributions on feasible paths .…

The concept of machine biometrics is proposed in this work for the first time, as a way to authenticate machine identities interacting with humans in everyday life . The experimental results revealed the ability of the proposedbiometrics to identify cars with high accuracy up to 98% for the case of the .…

We study the problem of maximizing the geometric mean of $d$ low-degreenon-negative forms on the real or complex sphere in $n$ variables . We show that this highly non-convex problem is NP-hard even when the forms are quadratic and is equivalent to optimizing a homogeneous polynomial of degree $O(d) on thesphere .…

This paper presents an Alzheimer’s disease detection method based on learning structural similarity between Magnetic Resonance Images (MRIs) and representing this similarity as a graph . Methods: We construct thesimilarity graph using embedded features of the input image (i.e., Non-Demented(ND), Very Mild Demented (VMD), Mild Dementsed (MD) and Moderated (MDTD) We use the similarity graph as aregularizer in the loss function of a CNN model to minimize the distance between the input images and their k-nearest neighbours .…

ANEA is a tool to automatically annotate named entities in text based on entity lists . It spans the whole pipeline from obtaining the lists to analyzing the errors of the distant supervision . A tuning step allows the userto improve the automatic annotation with their linguistic insights withouthaving to manually label or check all tokens .…

A popular assumption for out-of-distribution generalization is that the training data comprises sub-datasets, each drawn from a distinct distribution . The goal is to “interpolate” these distributions and “extrapolate” beyond them — this objective is broadly known as domain generalization .…

Automated Essay Scoring (AES) is a cross-disciplinary effort involving education, Linguistics, and Natural Language Processing (NLP) The efficacy of an NLP model in AES tests it ability to evaluate long-term dependencies and extract meaning even when text is poorly written .…

PharmKE is a text analysis platform focused on thepharmaceutical domain . It performs text classification using state-of-the-art transfer learning models . The recognized entities are alsoused to expand the knowledge graph generated by DBpedia Spotlight for a given pharmaceutical text .…

Adversarial examples have appeared as a ubiquitous property of machinelearning models where bounded adversarial perturbation could mislead the modelsto make arbitrarily incorrect predictions . Such examples provide a way toassess the robustness of machine learning models as a proxy for understanding the model training process .…

We propose anovel query-efficient NonLinear Gradient Projection-based Boundary BlackboxAttack (NonLinear-BA) The code is publiclyavailable at https://://://github.com/AI-secure/NonLinearly-BA . We show that the projection-based boundary blackbox attacks are able to achieve muchsmaller magnitude of perturbations with 100% attack success rate based one efficient queries .…

Emotional images from sources of misinformation can greatly influence ourjudgements . Users are more likely to find sources as less credible and their content as biased . When sources portray specific politicians as angry, users find them less credible . These results highlight how implicit visual propositions manifested by emotions infacial expressions might have a substantial effect on our trust of news content and sources.…

Process mining is widely used to diagnose processes and uncover performance problems . We use results from causal inference and adapts these to be able to reason over event logs and process interventions . Our ProM plug-in produces recommendations that indicate how specific cases could have been handled differently to avoid a performance or compliance problem .…

We document an ongoing research process towards the implementation andintegration of a digital artefact, executing the lifecycle of a leveraged tradewith permissionless blockchain technology . By employing core functions of the’Dai Stablecoin system’ deployed on the . Ethereum blockchain, we produce theequivalent exposure of a .…

Galaxy clusters identified from the Sunyaev Zel’dovich (SZ) effect are a keyingredient in multi-wavelength cluster-based cosmology . We present a comparisonbetween two methods of cluster identification: the standard Matched Filter (MF)method in SZ cluster finding and a method using Convolutional Neural Networks(CNN) The CNN method requires very littlepre-processing of images, while the MF method requires little pre-processing .…

Deep learning approaches have shown great success in image classification tasks and can aid greatly towards the fast and reliable classification of pollen grain aerial imagery . The proposed approachearned a fourth-place in the final rankings in the ICPR-2020 Pollen Grain Classification Challenge; with a 0.972578 weighted F1 score, 0.950828 macroaverage F1 scores, and .972877 recognition accuracy .…

AwkwardForth is a Forth-driven virtual machine for deserializing data into Awkward Arrays . As a language, it is not intended for humans to write, but it loosens the coupling between Uproot and Awkward Array . It is about as fast as C++ ROOT and Avro and 10-80$\times$ faster thanfastavro.…

Experimental demonstration of complex robotic behaviors relies heavily on finding the correct controller gains . This painstaking process is often completed by a domain expert, requiring deep knowledge of the relationship between parameter values and the resulting behavior of the system .…

In conventional communication systems, any interference between twocommunicating points is regarded as unwanted noise since it distorts thereceived signals . On the other hand, allowing simultaneous transmission and intentionally accepting the interference of signals and even benefiting from it has been considered for a range of wireless applications .…

We extend the AAA (Adaptive-Antoulas-Anderson) algorithm to develop adata-driven modeling framework for linear systems with quadratic output (LQO) We show that by interpolating the transfer functionvalues on a subset of samples together with imposing a least-squaresminimization on the rest of the rest, we construct reliable data-driven LQO models .…

Municipal solid waste management is a major challenge for nowadays urbansocieties, because it accounts for a large proportion of public budget and,when mishandled, it can lead to environmental and social problems . This article contributes with an exact multiobjective approach to solve the waste bin location in which the optimization criteria that are considered are: the accessibility to the system (as quality of servicemeasure), the investment cost, and the required frequency of waste removal from the bins .…

Accurate simulations of flows in stellar interiors are crucial to improving understanding of stellar structure and evolution . We demonstrate how discretization errors on grids ofreasonable size can lead to spurious flows orders of magnitude faster than thephysical flow . Well-balanced numerical schemes can deal with this problem .…

The aim of this work is to study solitary water wave interactions between twotopographic obstacles for the forced Korteweg-de Vries equation (fKdV) We identify regimes in which solitary wave interactions maintain two well separated crests and regimes where the number of local maxima varies according to the laws $2/2/1/2$ or $2\rightarrow .…

Abacus is an $N$-body simulation codethat allocates all persistent state in POSIX shared memory, or ramdisk . The main simulation executable is invoked once per time step, memorymapping the input state, computing the output state directly into ramdisk, and unmapping it.…

This paper studies the application of the blended dynamics approach towards the distributed optimization problem . The benefits include (i) individual cost functionneed not be convex as long as the global cost function is strongly convex and (ii) the convergence rate of the distributed algorithm is arbitrarily close to the convergence rates of the centralized one .…

We study the problem of community recovery from coarse measurements of agraph . We build on thestochastic block model by mathematically formalizing the coarsening process . We characterize an error bound for communityrecovery . The error bound yields simple and closed-form asymptotic conditionsto achieve the perfect recovery of the coarse graph communities, the authors say .…

This survey analyses the role of data-driven methodologies for pandemicmodelling and control . We provide a roadmap from the access to epidemiological data sources to the control of epidemic phenomena . A 3M-analysis is presented, whose three pillars are: Monitoring,Modelling and Managing .…

Many manipulation tasks can be cast as a sequence of spatial relationships and constraints between objects . We aim to discover and scalethese task-specific spatial relationships by representing manipulation tasks asoperations over graphs . To do this, we pose manipulating a large, variablenumber of objects as a probabilistic classification problem .…

In this work, we present a per-instant pose optimization method that can generate configurations that achieve specified pose or motion objectives asbest as possible over a sequence of solutions . We demonstrate how to effectively incorporate environment collisionavoidance as a single term in this multi-objective, optimization-based IKstructure .…

VPIC is a general purpose Particle-in-Cell simulation code for modeling plasma phenomena such as magnetic reconnection, fusion, solar weather, and laser-plasma interaction in three dimensions . VPIC’s capacity in both fidelity and scale makes it particularly well-suited for plasma research on pre-exascale and exascale platforms .…

In this work, we design and investigate contrast-independent partiallyexplicit time discretizations for wave equations in heterogeneous high-contrast media . We consider multiscale problems, where the spatial heterogeneities are not resolved . The splitting requires a careful design. We prove that the proposed splitting isunconditionally stable under some suitable conditions formulated for the secondspace (slow) We present numerical results and show that proposed methods provide results similar to implicit methods with the time step that is independent of the contrast.…

Current Cloud solutions for Edge Computing are inefficient for data-centric applications, as they focus on the IaaS/PaaS level and miss the datamodeling and operations perspective . MEDAL is an intelligent Cloud-to-Edge Data Fabric to support Data Operations(DataOps)across the continuum and to automate management and orchestrationoperations over a combined view of the data and the resource layer .…

Everyedge $e in G-T$ induces a cycle in $T $T \cup \{e\}$. The intersection of twodistinct such cycles is the set of edges of $T$ that belong to both cycles . Weconsider the problem of finding a spanning tree that has the least number of non-empty intersections .…