Search

State-of-the-art face restoration methods employ deep convolutional neural networks (CNNs) to learn a mapping between degraded and sharp facial patterns by exploring local appearance knowledge. However, most of these methods do not well exploit facial structures and identity information, and only deal with task-specific face restoration (e.g., face super-resolution or deblurring). In this paper, we propose cross-tasks and cross-models plug-and-play 3D facial priors to explicitly embed the network with the sharp facial structures for general face restoration tasks. Our 3D priors are the first to explore 3D morphable knowledge based on the fusion of parametric descriptions of face attributes (e.g., identity, facial expression, texture, illumination, and face pose). Furthermore, the priors can easily be incorporated into any network and are very efficient in improving the performance and accelerating the convergence speed. Firstly, a 3D face rendering branch is set up to obtain 3D priors of salient facial structures and identity knowledge. Secondly, for better exploiting this hierarchical information (i.e., intensity similarity, 3D facial structure, and identity content), a spatial attention module is designed for the image restoration problems. Extensive face restoration experiments including face super-resolution and deblurring demonstrate that the proposed 3D priors achieve superior face restoration results over the state-of-the-art algorithms.

Wide-scale adoption of optoacoustic imaging in biology and medicine critically depends on availability of affordable scanners combining ease of operation with optimal imaging performance. Here we introduce LightSpeed: a low-cost real-time volumetric handheld optoacoustic imager based on a new compact software-defined ultrasound digital acquisition platform and a pulsed laser diode. It supports the simultaneous signal acquisition from up to 192 ultrasound channels and provides a hig-bandwidth direct optical link (2x 100G Ethernet) to the host-PC for ultra-high frame rate image acquisitions. We demonstrate use of the system for ultrafast (500Hz) 3D human angiography with a rapidly moving handheld probe. LightSpeed attained image quality comparable with a conventional optoacoustic imaging systems employing bulky acquisition electronics and a Q-switched pulsed laser. Our results thus pave the way towards a new generation of compact, affordable and high-performance optoacoustic scanners.

Context: Code review is a fundamental, yet expensive part of software engineering. Therefore, research on understanding code review and its efficiency and performance is paramount. Objective: We aim to test the effect of a guidance approach on review effectiveness and efficiency. This effect is expected to work by lowering the cognitive load of the task; thus, we analyze the mediation relationship as well. Method: To investigate this effect, we employ an experimental design where professional developers have to perform three code reviews. We use three conditions: no guidance, a checklist, and a checklist-based review strategy. Furthermore, we measure the reviewers' cognitive load. Limitations: The main limitations of this study concern the specific cohort of participants, the mono-operation bias for the guidance conditions, and the generalizability to other changes and defects. Full registered report: https://doi.org/10.17605/OSF.IO/5FPTJ; Materials: https://doi.org/10.6084/m9.figshare.11806656

Semantic segmentation of 3D point clouds is a crucial task in scene understanding and is also fundamental to indoor scene applications such as indoor navigation, mobile robotics, augmented reality. Recently, deep learning frameworks have been successfully adopted to point clouds but are limited by the size of data. While most existing works focus on individual sampling points, we use surface patches as a more efficient representation and propose a novel indoor scene segmentation framework called patch graph convolution network (PGCNet). This framework treats patches as input graph nodes and subsequently aggregates neighboring node features by dynamic graph U-Net (DGU) module, which consists of dynamic edge convolution operation inside U-shaped encoder–decoder architecture. The DGU module dynamically update graph structures at each level to encode hierarchical edge features. Incorporating PGCNet, we can segment the input scene into two types, i.e., room layout and indoor objects, which is afterward utilized to carry out final rich semantic labeling of various indoor scenes. With considerable speedup training, the proposed framework achieves effective performance equivalent to state-of-the-art for segmenting standard indoor scene dataset.

The second workshop on intelligent healthy interfaces (HEALTHI), collocated with the 2022 ACM Intelligent User Interfaces (IUI) conference, offers a forum that brings academics and industry researchers together and seeks submissions broadly related to the design of healthy user interfaces. The workshop will discuss intelligent user interfaces such as screens, wearables, voices assistants, and chatbots in the context of accessibly supporting health, health behavior, and wellbeing.

In the last years, the number of connections in mobile telecommunication networks has increased rampantly, and in consequence, the number and type of relationships among entities. Should such interactions are to be profitable, entities will need to rely on each other. Hence, mobile telecommunication networks demand trust and reputation models that allow developing feasible communications in 5G and beyond networks, through which a group of entities can establish chains of services between cross-operators/domains, with security and trustworthiness. One of the key obstacles to achieving generalized connectivity beyond 5G networks is the lack of automatized, efficient, and scalable models for establishing security and trust. In this vein, this article proposes a pre-standardization approach for reputation-based trust models beyond 5G. To this end, we have realized a thorough review of the literature to match trust standardization approaches. An abstract set of requirements and key performance indicators has been extracted, and some pre-standardization recommendations proposed to fulfill essential conditions of future networks and to cover the lack of common trust and reputation models beyond 5G.

Computing path queries such as the shortest path in public transport networks is challenging because the path costs between nodes change over time. A reachability query from a node at a given start time on such a network retrieves all points of interest (POIs) that are reachable within a given cost budget. Reachability queries are essential building blocks in many applications, for example, group recommendations, ranking spatial queries, or geomarketing. We propose an efficient solution for reachability queries in public transport networks. Currently, there are two options to solve reachability queries. (1) Execute a modified version of Dijkstra’s algorithm that supports time-dependent edge traversal costs; this solution is slow since it must expand edge by edge and does not use an index. (2) Issue a separate path query for each single POI, i.e., a single reachability query requires answering many path queries. None of these solutions scales to large networks with many POIs. We propose a novel and lightweight reachability index. The key idea is to partition the network into cells. Then, in contrast to other approaches, we expand the network cell by cell. Empirical evaluations on synthetic and real-world networks confirm the efficiency and the effectiveness of our index-based reachability query solution.

Most of the current Brain–Computer Interfaces (BCIs) application scenarios use electroencephalographic signals (EEG) containing the subject’s information. It means that if EEG were maliciously manipulated, the proper functioning of BCI frameworks could be at risk. Unfortunately, it happens in frameworks sensitive to noise-based cyberattacks, and more efforts are needed to measure the impact of these attacks. This work presents and analyzes the impact of four noise-based cyberattacks attempting to generate fake P300 waves in two different phases of a BCI framework. A set of experiments show that the greater the attacker’s knowledge regarding the P300 waves, processes, and data of the BCI framework, the higher the attack impact. In this sense, the attacker with less knowledge impacts 1% in the acquisition phase and 4% in the processing phase, while the attacker with the most knowledge impacts 22% and 74%, respectively.

Updates on ontologies affect the operations built on top of them. But not all changes are equal: some updates drastically change the result of operations; others lead to minor variations, if any. Hence, estimating the impact of a change ex-ante is highly important, as it might make ontology engineers aware of the consequences of their action during editing. However, in order to estimate the impact of changes, we need to understand how to measure them.To address this gap for embeddings, we propose a new measure called Embedding Resemblance Indicator (ERI), which takes into account both the stochasticity of learning embeddings as well as the shortcomings of established comparison methods. We base ERI on (i) a similarity score, (ii) a robustness factor $\hatμ $ (based on the embedding method, similarity measure, and dataset), and (iii) the number of added or deleted entities to the embedding computed with the Jaccard index.To evaluate ERI, we investigate its usage in the context of two biomedical ontologies and three embedding methods---GraRep, LINE, and DeepWalk---as well as the two standard benchmark datasets---FB15k-237 and Wordnet-18-RR---with TransE and RESCAL embeddings. To study different aspects of ERI, we introduce synthetic changes in the knowledge graphs, generating two test-cases with five versions each and compare their impact with the expected behaviour. Our studies suggests that ERI behaves as expected and captures the similarity of embeddings based on the severity of changes. ERI is crucial for enabling further studies into impact of changes on embeddings.

Segmentation of anatomical structures is a fundamental image analysis task for many applications in the medical field. Deep learning methods have been shown to perform well, but for this purpose large numbers of manual annotations are needed in the first place, which necessitate prohibitive levels of resources that are often unavailable. In an active learning framework of selecting informed samples for manual labeling, expert clinician time for manual annotation can be optimally utilized, enabling the establishment of large labeled datasets for machine learning. In this paper, we propose a novel method that combines representativeness with uncertainty in order to estimate ideal samples to be annotated, iteratively from a given dataset. Our novel representativeness metric is based on Bayesian sampling, by using information-maximizing autoencoders. We conduct experiments on a shoulder magnetic resonance imaging (MRI) dataset for the segmentation of four musculoskeletal tissue classes. Quantitative results show that the annotation of representative samples selected by our proposed querying method yields an improved segmentation performance at each active learning iteration, compared to a baseline method that also employs uncertainty and representativeness metrics. For instance, with only 10% of the dataset annotated, our method reaches within 5% of Dice score expected from the upper bound scenario of all the dataset given as annotated (an impractical scenario due to resource constraints), and this gap drops down to a mere 2% when less than a fifth of the dataset samples are annotated. Such active learning approach to selecting samples to annotate enables an optimal use of the expert clinician time, being often the bottleneck in realizing machine learning solutions in medicine., Comment: 10 pages, 7 figures

While supervised object detection and segmentation methods achieve impressive accuracy, they generalize poorly to images whose appearance significantly differs from the data they have been trained on. To address this when annotating data is prohibitively expensive, we introduce a self-supervised detection and segmentation approach that can work with single images captured by a potentially moving camera. At the heart of our approach lies the observation that object segmentation and background reconstruction are linked tasks, and that, for structured scenes, background regions can be re-synthesized from their surroundings, whereas regions depicting the moving object cannot. We encode this intuition into a self-supervised loss function that we exploit to train a proposal-based segmentation network. To account for the discrete nature of the proposals, we developed a Monte Carlo-based training strategy that allows the algorithm to explore the large space of object proposals. We apply our method to human detection and segmentation in images that visually depart from those of standard benchmarks and outperform existing self-supervised methods.

Item does not contain fulltext Scene parsing is a critical component for understanding complex scenes in applications such as autonomous driving. Semantic segmentation networks are typically reported for scene parsing but semantic edge networks have also become of interest because of the sparseness of the segmented maps. This work presents an end-to-end trained lightweight deep semantic edge detection architecture called LiteEdge suitable for edge deployment. By utilizing hierarchical supervision and a new weighted multi-label loss function to balance different edge classes during training, LiteEdge predicts with high accuracy category-wise binary edges. Our LiteEdge network with only approximately 3M parameters, has a semantic edge prediction accuracy of 52.9% mean maximum F (MF) score on the Cityscapes dataset. This accuracy was evaluated on the network trained to produce a low resolution edge map. The network can be quantized to 6-bit weights and 8-bit activations and shows only a 2% drop in the mean MF score. This quantization leads to a memory footprint savings of 6X for an edge device. ICCV 2021: The IEEE/CVF International Conference on Computer Vision (11-17 October, 2021)

As Java 8 introduced functional interfaces and lambda expressions to the Java programming language, the JDK API was changed to introduce support for lambda expressions, thus allowing consumers to define lambda functions when using Java’s collections. While the JDK API allows for a functional paradigm, for API consumers to be able to completely embrace Java’s new functional features, third-party APIs must also support lambda expressions. To understand the current state of the Java ecosystem, we investigate (i) the extent to which third-party Java APIs have changed their interfaces, (ii) why or why not they introduce functional interface support and (iii) in the case the API has changed its interface how it does so. We also investigate the consumers’ perspective, particularly their ease in using lambda expressions in Java with APIs. We perform our investigation by manually analyzing the top 50 popular Java APIs, conducting in-person and email interviews with 23 API producers, and surveying 110 developers. We find that only a minority of the top 50 APIs support functional interfaces, the rest does not support them, predominantly in the interest of backward compatibility. Java 7 support is still greatly desirable due to enterprise projects not migrating to newer versions of Java. This suggests that the Java ecosystem is stagnant and that the introduction of new language features will not be enough to save it from the advent of new languages such as Kotlin (JVM based) and Rust (non-JVM based).

Regression testing comprises techniques which are applied during software evolution to uncover faults effectively and efficiently. While regression testing is widely studied for functional tests, performance regression testing, e.g., with software microbenchmarks, is hardly investigated. Applying test case prioritization (TCP), a regression testing technique, to software microbenchmarks may help capturing large performance regressions sooner upon new versions. This may especially be beneficial for microbenchmark suites, because they take considerably longer to execute than unit test suites. However, it is unclear whether traditional unit testing TCP techniques work equally well for software microbenchmarks. In this paper, we empirically study coverage-based TCP techniques, employing total and additional greedy strategies, applied to software microbenchmarks along multiple parameterization dimensions, leading to 54 unique technique instantiations. We find that TCP techniques have a mean APFD-P (average percentage of fault-detection on performance) effectiveness between 0.54 and 0.71 and are able to capture the three largest performance changes after executing 29% to 66% of the whole microbenchmark suite. Our efficiency analysis reveals that the runtime overhead of TCP varies considerably depending on the exact parameterization. The most effective technique has an overhead of 11% of the total microbenchmark suite execution time, making TCP a viable option for performance regression testing. The results demonstrate that the total strategy is superior to the additional strategy. Finally, dynamic-coverage techniques should be favored over static-coverage techniques due to their acceptable analysis overhead; however, in settings where the time for prioritzation is limited, static-coverage techniques provide an attractive alternative.

The multi-modal and interrelated nature of lifelog data makes it well suited for graph-based representations. In this paper, we present the second iteration of LifeGraph, a Knowledge Graph for Lifelog Data, initially introduced during the 3rd Lifelog Search Challenge in 2020. This second iteration incorporates several lessons learned from the previous version. While the actual graph has undergone only small changes, the mechanisms by which it is traversed during querying as well as the underlying storage system which performs the traversal have been changed. The means for query formulation have also been slightly extended in capability and made more efficient and intuitive. All these changes have the aim of improving result quality and reducing query time.

Neuromorphic chips that use spikes to encode information could provide fast and energy-efficient computing for ubiquitous embedded systems. A bio-plausible spike-timing solution for training spiking neural networks that makes the most of sparsity is implemented on the BrainScaleS-2 hardware platform.

Methods for processing point cloud information have seen a great success in collider physics applications. One recent breakthrough in machine learning is the usage of Transformer networks to learn semantic relationships between sequences in language processing. In this work, we apply a modified Transformer network called Point Cloud Transformer as a method to incorporate the advantages of the Transformer architecture to an unordered set of particles resulting from collision events. To compare the performance with other strategies, we study jet-tagging applications for highly-boosted particles., Comment: 12 pages, 3 figures

Code review is a widely adopted software development practice aimed at finding defects, improving software quality, and transferring knowledge among developers. Performing an effective code review is a challenging task for developers. Two of the main challenges reviewers face are (1) understanding the content of a review change-set and (2) assessing the impact of a change on the codebase. Visualization techniques can be used to increase developers’ understanding of a changeset to review and its context. However, only a few attempts have been made to apply visualization to code review.In this paper, we present a novel approach we devised to support developers in understanding GitHub pull requests. Our approach expands the GitHub interface with two lateral bars to let developers navigate to the definition/uses of the methods in the changeset under review.We evaluated our approach’s interface through (1) interviews with eight developers and (2) a survey with 12 participants. Based on the results of this evaluation, we implemented our approach in a web-based tool, ChangeViz.Pre-print, data and materials, and demo video: https://doi.org/10.5281/zenodo.5175927.

System dynamics (SD) is an established discipline to model and simulate complex dynamic systems. The primary goal of SD is to evaluate and design new policies that can impact the system under study in a desired way. Policy design, that is, identifying effective model levers, however, is a challenge and in many cases trial-and-error driven. In this article, we introduce a new and coherent framework for model analysis, called structural analysis methods (SAM), to facilitate the policy design process in complex SD models. SAM provides a resource-efficient and effective means for the detection of candidate policy parameters. It enables to identify intended and unintended effects of activating these policy parameters, and to discover candidate structural changes such as introducing new variables and links in SD models. The main innovation of SAM is that it translates the structure of SD models into weighted digraphs allowing algorithmic tools from the realms of graph theory and network science to be applied to SD. SAM is validated on the basis of two well-known simulation models of increasing complexity: the third-order Phosphorus Loops in Soil and Sediment (PLUM) model and the fifth-order World2 model. The validation shows that SAM seems to be most valuable for the analysis of more complex simulation models (World2) and is less suited for the analysis of low complexity models (PLUM)., Simulation Modelling Practice and Theory, 110, ISSN:1569-190X

Data analysis in high-dimensional spaces aims at obtaining a synthetic description of a data set, revealing its main structure and its salient features. We here introduce an approach providing this description in the form of a topography of the data, namely a human-readable chart of the probability density from which the data are harvested. The approach is based on an unsupervised extension of Density Peak clustering and a non-parametric density estimator that measures the probability density in the manifold containing the data. This allows finding automatically the number and the height of the peaks of the probability density, and the depth of the "valleys" separating them. Importantly, the density estimator provides a measure of the error, which allows distinguishing genuine density peaks from density fluctuations due to finite sampling. The approach thus provides robust and visual information about the density peaks' height, their statistical reliability, and their hierarchical organization, offering a conceptually powerful extension of the standard clustering partitions. We show that this framework is particularly useful in the analysis of complex data sets., There is a Supplementary Information document in the ancillary files folder

Continuous integration is an agile software development practice. Instead of integrating features right before a release, they are constantly being integrated into an automated build process. This shortens the release cycle, improves software quality, and reduces time to market. However, the whole process will come to a halt when a commit breaks the build, which can happen for several reasons, e.g., compilation errors or test failures, and fixing the build suddenly becomes a top priority. Developers not only have to find the cause of the build break and fix it, but they have to be quick in all of it to avoid a delay for others. Unfortunately, these steps require deep knowledge and are often time-consuming. To support developers in fixing a build break, we propose Bart, a tool that summarizes the reasons for Maven build failures and suggests possible solutions found on the internet. We will show in a case study with 17 participants that developers find Bart useful to understand build breaks and that using Bart substantially reduces the time to fix a build break, on average by 37%. We have also conducted a qualitative study to better understand the workflows and information needs when fixing builds. We found that typical workflows differ substantially between various error categories, and that several uncommon build errors are both very hard to investigate and to fix. These findings will be useful to inform future research in this area.

Refactoring is a well-established practice that aims at improving the internal structure of a software system without changing its external behavior. Existing literature provides evidence of how and why developers perform refactoring in practice. In this paper, we continue on this line of research by performing a large-scale empirical analysis of refactoring practices in 200 open source systems. Specifically, we analyze the change history of these systems at commit level to investigate: (i) whether developers perform refactoring operations and, if so, which are more diffused and (ii) when refactoring operations are applied, and (iii) which are the main developer-oriented factors leading to refactoring. Based on our results, future research can focus on enabling automatic support for less frequent refactorings and on recommending refactorings based on the developer's workload, project's maturity and developer's commitment to the project.

Algorithms for Magnetic Resonance (MR) image reconstruction from undersampled measurements exploit prior information to compensate for missing k-space data. Deep learning (DL) provides a powerful framework for extracting such information from existing image datasets, through learning, and then using it for reconstruction. Leveraging this, recent methods employed DL to learn mappings from undersampled to fully sampled images using paired datasets, including undersampled and corresponding fully sampled images, integrating prior knowledge implicitly. In this article, we propose an alternative approach that learns the probability distribution of fully sampled MR images using unsupervised DL, specifically Variational Autoencoders (VAE), and use this as an explicit prior term in reconstruction, completely decoupling the encoding operation from the prior. The resulting reconstruction algorithm enjoys a powerful image prior to compensate for missing k-space data without requiring paired datasets for training nor being prone to associated sensitivities, such as deviations in undersampling patterns used in training and test time or coil settings. We evaluated the proposed method with T1 weighted images from a publicly available dataset, multicoil complex images acquired from healthy volunteers (N=8) and images with white matter lesions. The proposed algorithm, using the VAE prior, produced visually high quality reconstructions and achieved low RMSE values, outperforming most of the alternative methods on the same dataset. On multi-coil complex data, the algorithm yielded accurate magnitude and phase reconstruction results. In the experiments on images with white matter lesions, the method faithfully reconstructed the lesions., IEEE Transactions on Medical Imaging, 38 (7), ISSN:0278-0062, ISSN:1558-254X

Change-based code review is a software quality assurance technique that is widely used in practice. Therefore, better understanding what influences performance in code review and finding ways to improve it can have a large impact. In this study, we examine the association of working memory capacity and cognitive load to code review performance and we test the predictions of a recent theory regarding improved code review efficiency with certain code change part orders. We perform a confirmatory experiment with 50 participants, mostly professional software developers. The participants performed code reviews on one small and two larger code changes from an open source software system to which we had seeded additional defects. We measured their efficiency and effectiveness in defect detection, their working memory capacity, and several potential confounding factors. We find that there is a moderate association between working memory capacity and the effectiveness of finding delocalized defects, influenced by other factors, whereas the association with other defect types is almost non-existing. We also confirm that the effectiveness of reviews is significantly larger for small code changes. We observe a tendency that the order of presenting the code changes influences the efficiency of code review, but cannot conclude this reliably., This is a pre-print of an article published in Empirical Software Engineering. The final authenticated version is available online at: https://doi.org/10.1007/s10664-018-9676-8

There is a growing use of intelligent systems to support human decision-making across several domains. Trust in intelligent systems, however, is pivotal in shaping their widespread adoption. Little is currently understood about how trust in an intelligent system evolves over time and how it is mediated by the accuracy of the system. We aim to address this knowledge gap by exploring trust formation over time and its relation to system accuracy. To that end, we built an intelligent house recommendation system and carried out a longitudinal study consisting of 201 participants across 3 sessions in a week. In each session, participants were tasked with finding housing that fit a given set of constraints using a conventional web interface that reflected a typical housing search website. Participants could choose to use an intelligent decision support system to help them find the right house. Depending on the group, participants received a variation of accurate or inaccurate advice from the intelligent system throughout each session. We measured trust using a trust in automation scale at the end of each session. We found evidence suggesting that trust development is a slow process that evolves over multiple sessions, and that first impressions of the intelligent system are highly influential. Our results echo earlier research on trust formation in single session interactions, corroborating that reliability, validity, predictability, and dependability all influence trust formation. We also found that the age of the participants and their affinity with technology had an effect on their trust in the intelligent system. Our findings highlight the importance of first impressions and improvement of system accuracy for trust development. Hence, our study is an important first step in understanding trust development, breakdown of trust, and trust repair over multiple system interactions, informing improved system design.

In the research of video retrieval systems, comparative assessments during dedicated retrieval competitions provide priceless insights into the performance of individual systems. The scope and depth of such evaluations is unfortunately hard to improve, due to the limitations by the set-up costs, logistics and organization complexity of large events. We show that this easily impairs the statistical significance of the collected results, and the reproducibility of the competition outcomes. In this paper, we present a methodology for remote comparative evaluations of content-based video retrieval systems and demonstrate that such evaluations scale-up to sizes that reliably produce statistically robust results, and propose additional measures that increase the replicability of the experiment. The proposed remote evaluation methodology forms a major contribution towards open science in interactive retrieval benchmarks. At the same time, the detailed evaluation reports form an interesting source of new observations about many subtle, previously inaccessible aspects of video retrieval.

In radio astronomy, antennas monitor portions of the sky to collect radio signals. The antennas produce data streams that are of high volume and velocity (~2.5 GB/s) and the inverse Fourier transform is used to convert the collected signals into sky images that astrophysicists use to conduct their research. Applying the inverse Fourier transform in a streaming setting, however, is not ideal since its computational complexity is quadratic in the size of the image.In this article, we propose the Single Point Incremental Fourier Transform (SPIFT), a novel incremental algorithm to produce sequences of sky images. SPIFT computes the Fourier transform for a new signal in a linear number of complex multiplications by exploiting twiddle factors, multiplicative constant coefficients. We prove that twiddle factors are periodic and show how circular shifts can be exploited to reuse multiplication results. The cost of the additive operations can be curbed by exploiting the embarrassingly parallel nature of the additions, which modern big data streaming frameworks can leverage to compute slices of the image in parallel. Our experiments suggest that SPIFT can efficiently generate sequences of sky images: it computes the complex multiplications 4 to 12x faster than the Discrete Fourier Transform, and its parallelisation of the additive operations shows linear speedup.

Mixed-initiative visual data analysis processes are characterized by the co-adaptation of users and systems over time. As the analysis progresses, both actors – users and systems – gather information, update their analysis behavior, and work on different tasks towards their respective goals. In this paper, we contribute a multigranular model of co-adaptive visual analysis that is centered around incremental learning goals derived from a hierarchical taxonomy of learning goals from pedagogy. Our model captures how both actors adapt their data-, task-, and user/system-models over time. We characterize interaction patterns in terms of the dynamics of learning and teaching that drive adaptation. To demonstrate our model’s applicability, we outline aspects of co-adaptation in related models of visual analytics and highlight co-adaptation in existing applications. We further postulate a set of expectations towards adaptation in mixed-initiative processes and identify open research questions and opportunities for future work in co-adaptation.

Item does not contain fulltext Deep Artificial Neural Networks (ANNs) employ a simplified analog neuron model that mimics the rate transfer function of integrate-and-fire neurons. In Spiking Neural Networks (SNNs), the predominant information transmission method is based on rate codes. This code is inefficient from a hardware perspective because the number of transmitted spikes is proportional to the encoded analog value. Alternate codes such as temporal codes that are based on single spikes are difficult to scale up for large networks due to their sensitivity to spike timing noise. Here we present a study of an encoding scheme based on temporal spike patterns. This scheme inherits the efficiency of temporal codes but retains the robustness of rate codes. The pattern code is evaluated on MNIST, CIFAR-10, and ImageNet image classification tasks. We compare the network performance of ANNs, rate-coded SNNs, and temporal-coded SNNs, using the classification error and operation count as performance metrics. We also estimate the power consumption of the digital logic needed for the operations associated with each encoding type, and the impact of the bit precision of the weights and activations. On ImageNet, the temporal pattern code achieves up to 35× reduction in the estimated power consumption compared to the rate-coded SNN, and 42× compared to the ANN. The classification error of the pattern-coded SNN is increased by

This paper presents an automatic speech translation system aimed at live subtitling of conference presentations. We describe the overall architecture and key processing components. More importantly, we explain our strategy for building a complex system for end-users from numerous individual components, each of which has been tested only in laboratory conditions. The system is a working prototype that is routinely tested in recognizing English, Czech, and German speech and presenting it translated simultaneously into 42 target languages.

Software benchmarks are only as good as the performance measurements they yield. Unstable benchmarks show high variability among repeated measurements, which causes uncertainty about the actual performance and complicates reliable change assessment. However, if a benchmark is stable or unstable only becomes evident after it has been executed and its results are available. In this paper, we introduce a machine-learning-based approach to predict a benchmark’s stability without having to execute it. Our approach relies on 58 statically-computed source code features, extracted for benchmark code and code called by a benchmark, related to (1) meta information, e.g., lines of code (LOC), (2) programming language elements, e.g., conditionals or loops, and (3) potentially performance-impacting standard library calls, e.g., file and network input/output (I/O). To assess our approach’s effectiveness, we perform a large-scale experiment on 4,461 Go benchmarks coming from 230 open-source software (OSS) projects. First, we assess the prediction performance of our machine learning models using 11 binary classification algorithms. We find that Random Forest performs best with good prediction performance from 0.79 to 0.90, and 0.43 to 0.68, in terms of AUC and MCC, respectively. Second, we perform feature importance analyses for individual features and feature categories. We find that 7 features related to meta-information, slice usage, nested loops, and synchronization application programming interfaces (APIs) are individually important for good predictions; and that the combination of all features of the called source code is paramount for our model, while the combination of features of the benchmark itself is less important. Our results show that although benchmark stability is affected by more than just the source code, we can effectively utilize machine learning models to predict whether a benchmark will be stable or not ahead of execution. This enables spending precious testing time on reliable benchmarks, supporting developers to identify unstable benchmarks during development, allowing unstable benchmarks to be repeated more often, estimating stability in scenarios where repeated benchmark execution is infeasible or impossible, and warning developers if new benchmarks or existing benchmarks executed in new environments will be unstable.

Executing software microbenchmarks, a form of small-scale performance tests predominantly used for libraries and frameworks, is a costly endeavor. Full benchmark suites take up to multiple hours or days to execute, rendering frequent checks, e.g., as part of continuous integration (CI), infeasible. However, altering benchmark configurations to reduce execution time without considering the impact on result quality can lead to benchmark results that are not representative of the software’s true performance. We propose the first technique to dynamically stop software microbenchmark executions when their results are sufficiently stable. Our approach implements three statistical stoppage criteria and is capable of reducing Java Microbenchmark Harness (JMH) suite execution times by 48.4% to 86.0%. At the same time it retains the same result quality for 78.8% to 87.6% of the benchmarks, compared to executing the suite for the default duration. The proposed approach does not require developers to manually craft custom benchmark configurations; instead, it provides automated mechanisms for dynamic reconfiguration. Hence, making dynamic reconfiguration highly effective and efficient, potentially paving the way to inclusion of JMH microbenchmarks in CI.

Businesses were moving during the past decades to-ward full digital models, which made companies face new threats and cyberattacks affecting their services and, consequently, their profits. To avoid negative impacts, companies' investments in cybersecurity are increasing considerably. However, Small and Medium-sized Enterprises (SMEs) operate on small budgets, minimal technical expertise, and few personnel to address cybersecurity threats. In order to address such challenges, it is essential to promote novel approaches that can intuitively present cybersecurity-related technical information.This paper introduces SecBot, a cybersecurity-driven conversational agent (i.e., chatbot) for the support of cybersecurity planning and management. SecBot applies concepts of neural networks and Natural Language Processing (NLP), to interact and extract information from a conversation. SecBot can (a) identify cyberattacks based on related symptoms, (b) indicate solutions and configurations according to business demands, and (c) provide insightful information for the decision on cybersecurity investments and risks. A formal description had been developed to describe states, transitions, a language, and a Proof-of-Concept (PoC) implementation. A case study and a performance evaluation were conducted to provide evidence of the proposed solution's feasibility and accuracy.

The recently developed optoacoustic tomography systems have attained volumetric frame rates exceeding 100 Hz, thus opening up new venues for studying previously invisible biological dynamics. Further gains in temporal resolution can potentially be achieved via partial data acquisition, though a priori knowledge on the acquired data is essential for rendering accurate reconstructions using compressed sensing approaches. In this work, we suggest a machine learning method based on principal component analysis for high-frame-rate volumetric cardiac imaging using only a few tomographic optoacoustic projections. The method is particularly effective for discerning periodic motion, as demonstrated herein by non-invasive imaging of a beating mouse heart. A training phase enables efficiently compressing the heart motion information, which is subsequently used as prior information for image reconstruction from sparse sampling at a higher frame rate. It is shown that image quality is preserved with a 64-fold reduction in the data flow. We demonstrate that, under certain conditions, the volumetric motion could effectively be captured by relying on time-resolved data from a single optoacoustic detector. Feasibility of capturing transient (non-periodic) events not registered in the training phase is further demonstrated by visualizing perfusion of a contrast agent in vivo. The suggested approach can be used to significantly boost the temporal resolution of optoacoustic imaging and facilitate development of more affordable and data efficient systems., IEEE Transactions on Medical Imaging, 39 (10), ISSN:0278-0062, ISSN:1558-254X

An effective and efficient application of Continuous Integration (CI) and Delivery (CD) requires software projects to follow certain principles and good practices. Configuring such a CI/CD pipeline is challenging and error-prone. Therefore, automated linters have been proposed to detect errors in the pipeline. While existing linters identify syntactic errors, detect security vulnerabilities or misuse of the features provided by build servers, they do not support developers that want to prevent common misconfigurations of a CD pipeline that potentially violate CD principles ("CD smells"). To this end, we propose CD-Linter, a semantic linter that can automatically identify four different smells in such a pipeline configuration file. We have evaluated our approach through a large-scale and long-term study that consists of (i) monitoring 145 issues (opened in as many open-source projects) over a period of 6 months, (ii) manually validating the detection precision and recall on a representative sample of issues, and (iii) assessing the magnitude of the observed smells on 5,312 open-source projects on GitLab. Our results show that CD smells are accepted and fixed by most of the developers and our linter achieves a precision of 87% and a recall of 94%. Those smells can be frequently observed in the wild, as 31% of projects with long configurations are affected by at least one smell. Data and Material, To appear in the proceedings of the ACM Joint European Software Engineering Conference and Symposium on the Foundations of Software Engineering (ESEC/FSE), Sun 8 - Fri 13 November 2020 Sacramento, California, United States.

Analytical queries often require a mixture of relational and linear algebra operations applied to the same data. This poses a challenge to analytic systems that must bridge the gap between relations and matrices. Previous work has mainly strived to fix the problem at the implementation level. This paper proposes a principled solution at the logical level. We introduce the relational matrix algebra (RMA), which seamlessly integrates linear algebra operations into the relational model and eliminates the dichotomy between matrices and relations. RMA is closed: All our relational matrix operations are performed on relations and result in relations; no additional data structure is required. Our implementation in MonetDB shows the feasibility of our approach, and empirical evaluations suggest that in-database analytics performs well for mixed workloads., Comment: 16 pages, 18 figures

Event cameras are novel sensors that output brightness changes in the form of a stream of asynchronous "events" instead of intensity frames. They offer significant advantages with respect to conventional cameras: high dynamic range (HDR), high temporal resolution, and no motion blur. Recently, novel learning approaches operating on event data have achieved impressive results. Yet, these methods require a large amount of event data for training, which is hardly available due the novelty of event sensors in computer vision research. In this paper, we present a method that addresses these needs by converting any existing video dataset recorded with conventional cameras to synthetic event data. This unlocks the use of a virtually unlimited number of existing video datasets for training networks designed for real event data. We evaluate our method on two relevant vision tasks, i.e., object recognition and semantic segmentation, and show that models trained on synthetic events have several benefits: (i) they generalize well to real event data, even in scenarios where standard-camera images are blurry or overexposed, by inheriting the outstanding properties of event cameras; (ii) they can be used for fine-tuning on real data to improve over state-of-the-art for both classification and semantic segmentation.

A Dark Pattern (DP) is an interface maliciously crafted to deceive users into performing actions they did not mean to do. In this work, we analyze Dark Patterns in 240 popular mobile apps and conduct an online experiment with 589 users on how they perceive Dark Patterns in such apps. The results of the analysis show that 95% of the analyzed apps contain one or more forms of Dark Patterns and, on average, popular applications include at least seven different types of deceiving interfaces. The online experiment shows that most users do not recognize Dark Patterns, but can perform better in recognizing malicious designs if informed on the issue. We discuss the impact of our work and what measures could be applied to alleviate the issue.

Visual Localization is one of the key enabling technologies for autonomous driving and augmented reality. High quality datasets with accurate 6 Degree-of-Freedom (DoF) reference poses are the foundation for benchmarking and improving existing methods. Traditionally, reference poses have been obtained via Structure-from-Motion (SfM). However, SfM itself relies on local features which are prone to fail when images were taken under different conditions, e.g., day/ night changes. At the same time, manually annotating feature correspondences is not scalable and potentially inaccurate. In this work, we propose a semi-automated approach to generate reference poses based on feature matching between renderings of a 3D model and real images via learned features. Given an initial pose estimate, our approach iteratively refines the pose based on feature matches against a rendering of the model from the current pose estimate. We significantly improve the nighttime reference poses of the popular Aachen Day-Night dataset, showing that state-of-the-art visual localization methods perform better (up to $47\%$) than predicted by the original reference poses. We extend the dataset with new nighttime test images, provide uncertainty estimates for our new reference poses, and introduce a new evaluation criterion. We will make our reference poses and our framework publicly available upon publication., 25 pages, 16 figures. Int J Comput Vis (2020)

In modern visual SLAM systems, it is a standard practice to retrieve potential candidate map points from overlapping keyframes for further feature matching or direct tracking. In this work, we argue that keyframes are not the optimal choice for this task, due to several inherent limitations, such as weak geometric reasoning and poor scalability. We propose a voxel-map representation to efficiently retrieve map points for visual SLAM. In particular, we organize the map points in a regular voxel grid. Visible points from a camera pose are queried by sampling the camera frustum in a raycasting manner, which can be done in constant time using an efficient voxel hashing method. Compared with keyframes, the retrieved points using our method are geometrically guaranteed to fall in the camera field-of-view, and occluded points can be identified and removed to a certain extend. This method also naturally scales up to large scenes and complicated multi-camera configurations. Experimental results show that our voxel map representation is as efficient as a keyframe map with 5 keyframes and provides significantly higher localization accuracy (average 46% improvement in RMSE) on the EuRoC dataset. The proposed voxel-map representation is a general approach to a fundamental functionality in visual SLAM and widely applicable.

In magnetic resonance imaging, the use of array detection and the number of detector elements have seen a steady increase over the past two decades. As a result, per-channel analog connection via long coaxial cable, as commonly used, poses an increasing challenge in terms of handling, safety, and coupling among cables. This situation is exacerbated when complementary recording of radiofrequency transmission or NMR-based magnetic field sensing further add to channel counts. A generic way of addressing this trend is the transition to digital signal transmission, enabled by digitization and first-level digital processing close to detector coils and sensors in the magnet bore. The foremost challenge that comes with this approach is to achieve high dynamic range, linearity, and phase stability despite interference by strong static, audiofrequency, and radiofrequency fields. The present work reports implementation of a 16-channel in-bore receiver, performing signal digitization and processing with subsequent optical transmission over fiber. Along with descriptions of the system design and construction, performance evaluation is reported. The resulting device is fully MRI compatible providing practically equal performance and signal quality compared to state-of-the-art RF digitizers operating outside the magnet. Its use is demonstrated by examples of head imaging and magnetic field recording.

There exist large amounts of numerical data that are stored in databases and must be analyzed. Database tables come with a schema and include non-numerical attributes; this is crucial contextual information that is needed for interpreting the numerical values. We propose relational matrix operations that support the analysis of data stored in tables and that preserve contextual information. The result of our approach are precisely defined relational matrix operations and a system implementation in MonetDB that illustrates the seamless integration of relational matrix operations into a relational DBMS.

Blockchains (BCs) are back-linked chain of records termed as blocks. To establish decentralized trusted systems, BCs employ consensus mechanisms. During the past ten years, there have been various proposals of BC design and implementations. However, most of the developed sate of the art BCs suffer from scalability issues. In order to enhance the scalability of the BCs, this paper proposes a transaction aggregation mechanism on a Proof-of-Stake (PoS)-based BC. Having developed the transaction aggregation and double linked blocks, efficient prevention and control of the BC's size growth is observed in the evaluated scenarios.

Spiking Neural Networks (SNNs) are bio-inspired networks that process information conveyed as temporal spikes rather than numeric values. An example of a sensor providing such data is the event-camera. It only produces an event when a pixel reports a significant brightness change. Similarly, the spiking neuron of an SNN only produces a spike whenever a significant number of spikes occur within a short period of time. Due to their spike-based computational model, SNNs can process output from event-based, asynchronous sensors without any pre-processing at extremely lower power unlike standard artificial neural networks. This is possible due to specialized neuromorphic hardware that implements the highly-parallelizable concept of SNNs in silicon. Yet, SNNs have not enjoyed the same rise of popularity as artificial neural networks. This not only stems from the fact that their input format is rather unconventional but also due to the challenges in training spiking networks. Despite their temporal nature and recent algorithmic advances, they have been mostly evaluated on classification problems. We propose, for the first time, a temporal regression problem of numerical values given events from an event-camera. We specifically investigate the prediction of the 3- DOF angular velocity of a rotating event-camera with an SNN. The difficulty of this problem arises from the prediction of angular velocities continuously in time directly from irregular, asynchronous event-based input. Directly utilising the output of event-cameras without any pre-processing ensures that we inherit all the benefits that they provide over conventional cameras. That is high-temporal resolution, high-dynamic range and no motion blur. To assess the performance of SNNs on this task, we introduce a synthetic event-camera dataset generated from real-world panoramic images and show that we can successfully train an SNN to perform angular velocity regression.

Adding more cameras to SLAM systems improves robustness and accuracy but complicates the design of the visual front-end significantly. Thus, most systems in the literature are tailored for specific camera configurations. In this work, we aim at an adaptive SLAM system that works for arbitrary multi-camera setups. To this end, we revisit several common building blocks in visual SLAM. In particular, we propose an adaptive initialization scheme, a sensor-agnostic, information- theoretic keyframe selection algorithm, and a scalable voxel- based map. These techniques make little assumption about the actual camera setups and prefer theoretically grounded methods over heuristics. We adapt a state-of-the-art visual- inertial odometry with these modifications, and experimental results show that the modified pipeline can adapt to a wide range of camera setups (e.g., 2 to 6 cameras in one experiment) without the need of sensor-specific modifications or tuning.

The mlt package implements maximum likelihood estimation in the class of conditional transformation models. Based on a suitable explicit parameterization of the unconditional or conditional transformation function using infrastructure from package basefun, we show how one can define, estimate, and compare a cascade of increasingly complex transformation models in the maximum likelihood framework. Models for the unconditional or conditional distribution function of any univariate response variable are set-up and estimated in the same computational framework simply by choosing an appropriate transformation function and parameterization thereof. As it is computationally cheap to evaluate the distribution function, models can be estimated by maximization of the exact likelihood, especially in the presence of random censoring or truncation. The relatively dense high-level implementation in the R system for statistical computing allows generalization of many established implementations of linear transformation models, such as the Cox model or other parametric models for the analysis of survival or ordered categorical data, to the more complex situations illustrated in this paper.

Source-code comments are an important communication medium between developers to better understand and maintain software. Current research focuses on auto-generating comments by summarizing the code. However, good comments contain additional details, like important design decisions or required trade-offs, and only developers can decide on the proper comment content. Automated summarization techniques cannot include information that does not exist in the code, therefore fully-automated approaches while helpful, will be of limited use. In our work, we propose to empower developers through a semi-automated system instead. We investigate the feasibility of using neural language models trained on a large corpus of Python documentation strings to generate completion suggestions and obtain promising results. By focusing on confident predictions, we can obtain a top-3 accuracy of over 70%, although this comes at the cost of lower suggestion frequency. Our models can be improved by leveraging context information like the signature and the full body of the method. Additionally, we are able to return good accuracy completions even for new projects, suggesting the generalizability of our approach.

One of the crucial problems in mathematical finance is to mitigate the risk of a financial position by setting up hedging positions of eligible financial securities. This leads to focusing on set-valued maps associating to any financial position the set of those eligible payoffs that reduce the risk of the position to a target acceptable level at the lowest possible cost. Among other properties of such maps, the ability to ensure lower semicontinuity and continuous selections is key from an operational perspective. It is known that lower semicontinuity generally fails in an infinite-dimensional setting. In this note we show that neither lower semicontinuity nor, more surprisingly, the existence of continuous selections can be a priori guaranteed even in a finite-dimensional setting. In particular, this failure is possible under arbitrage-free markets and convex risk measures., Comment: arXiv admin note: text overlap with arXiv:1702.01936

Authorship attribution (i.e., determining who is the author of a piece of source code) is an established research topic. State-of-the-art results for the authorship attribution problem look promising for the software engineering field, where they could be applied to detect plagiarized code and prevent legal issues. With this article, we first introduce a new language-agnostic approach to authorship attribution of source code. Then, we discuss limitations of existing synthetic datasets for authorship attribution, and propose a data collection approach that delivers datasets that better reflect aspects important for potential practical use in software engineering. Finally, we demonstrate that high accuracy of authorship attribution models on existing datasets drastically drops when they are evaluated on more realistic data. We outline next steps for the design and evaluation of authorship attribution models that could bring the research efforts closer to practical use for software engineering., 13 pages

Optoacoustic tomography systems have attained unprecedented volumetric imaging speeds, thus enabling insights into rapid biological dynamics and marking a milestone in the clinical translation of this modality. Fast imaging performance often comes at the cost of limited field-of-view, which may hinder potential applications looking at larger tissue volumes. The imaged field-of-view can potentially be expanded via scanning and using additional hardware to track the position of the imaging probe. However, this approach turns impractical for high-resolution volumetric scans performed in a freehand mode along arbitrary trajectories. We have developed an accurate framework for spatial compounding of time-lapse optoacoustic data. The method exploits the frequency-domain properties of vascular networks in optoacoustic images and estimates the relative motion and orientation of the imaging probe. This allows rapidly combining sequential volumetric frames into large area scans without additional tracking hardware. The approach is universally applicable for compounding volumetric data acquired with calibrated scanning systems but also in a freehand mode with up to six degrees of freedom. Robust performance is demonstrated for whole-body mouse imaging with spiral volumetric optoacoustic tomography and for freehand visualization of vascular networks in humans using volumetric imaging probes. The newly introduced capability for angiographic observations at multiple spatial and temporal scales is expected to greatly facilitate the use of optoacoustic imaging technology in pre-clinical research and clinical diagnostics. The technique can equally benefit other biomedical imaging modalities, such as scanning fluorescence microscopy, optical coherence tomography or ultrasonography, thus optimizing their trade-offs between fast imaging performance and field-of-view.