Multi-target tracking systems typically provide sets of estimated target states as their output. It is challenging to be able to integrate these outputs as inputs to other tracking systems to gain a better picture of the area under surveillance since they do not conform to the standard observation model. Moreover, in cyclic distributed systems, there may be common information between state estimates that would mean that fused estimates may become overconfident and corrupt the system. In this paper we develop a Bayesian multi-target estimator based on the covariance intersection algorithm for multi-target track-to-track data fusion. The approach is integrated into a multitarget tracking algorithm and demonstrated in simulations. The approach is able to account for missed tracks and false tracks produced by another tracking system.

Abstract—Continuum robots are increasingly used in a wide range of industrial applications due to their slender body and high redundancy, making them ideal for intervention in confined environments. However, most of the existing continuum robots operate with their backbone in a cantilever configuration, resulting in a limited accuracy at the tip because of their low stiffness. This paper proposes a new method to improve the stiffness of continuum robots, characterized by: (1) a modular section with adjustable stiffness (MSAS), able to inflate its diameter by ten times to provide physical support for a 6-DoF operating section; (2) a novel stiffness model developed to study the behaviour of the hybrid stiffening continuum arm; (3) a manual insertion method for an easy delivery with a limited number of actuators. The proposed stiffness model, which combines the stiffness characteristics of the MSAS and a 6-DoF continuum arm, predicts the performance of the novel hybrid continuum robot with an average deviation of 9.2% under external loads. A prototype has been demonstrated on a case study in a Trent 900 aeroengine, where the MSAS, with a natural diameter of 12.6 mm, is delivered through an inspection hole of 13 mm and then inflated (120 mm) to successfully lock the robot within the combustor (width: 200 mm; height: 110 mm). The physical support enables accurate operation with the 6-DoF continuum arm at the tip, indicating that the proposed design can improve the performance of conventional continuum robots.

This paper proposes an interpretable data-driven model Soft Lookup False data Matching (SLFM) through Dynamic Time Warping (DTW) for estimating missing/suspected SCADA data values. Numerical experiments conducted on one of the utilities of the Northern Region of the Indian Grid.

Exploiting the degree of freedom in the frequency domain and the near-far effect among different access points (APs), this paper proposes an opportunistic transmission scheme in cell-free massive MIMO-OFDM systems. The key idea is to orthogonally assign subcarriers among different users, so that there is only one user on each subcarrier. Then, a user is only served by its near APs through opportunistic selection, while the far APs are deactivated to avoid wasting power over their channels with severe propagation losses. Moreover, the number of active APs per subcarrier becomes small due to the opportunistic selection, making the use of downlink pilots and coherent detection feasible. As corroborated by numerical results, the proposed scheme can bring a significant performance boost in terms of both power efficiency and spectral efficiency.

The usability of computer vision is everywhere, whereas deep learning revolutionized the concept of artificial intelligence including computer vision. This paper discusses the leverages of deep learning for computer vision. At first, the background details of computer vision and deep learning have been discussed. Important tasks of computer vision like image classification, object localization, object detection, segmentation are briefly explained. Various types of deep learning algorithms have been described. The architecture of the convolutional neural network$ has been explained briefly. Applications of deep learning in different fields of computer vision like image classifications, human activity recognition, scene text detection and recognition, object tracking, visual question answering, etc., have been discussed with the references of the recent state-of-the-art works.

this paper comparing the the performance of different ITU-837 in predicting malaysias' rain attenuation CCDF.

This paper presents a literature review on Intention-aware and Interaction-aware Trajectory Prediction for Autonomous Vehicle, which covers primary studies since 2008. The research questions try to understand how intention/interaction improves prediction, and what are the techniques and technologies employed in trajectory prediction.

Brain signals are nonlinear and nonstationary time series, which provide information about spatiotemporal patterns of electrical activity in the brain. CHMMs are suitable tools for modeling multi-channel time-series dependent on both time and space, but state-space parameters grow exponentially with the number of channels. To cope with this limitation, we consider the influence model as the interaction of hidden Markov chains called Latent Structure Influence Models (LSIMs). LSIMs are capable of detecting nonlinearity and nonstationarity, making them well suited for multi-channel brain signals.

Stratified SGD (SSGD) is the primary approach for achieving serializable parallel SGD for matrix completion. State-of-the-art parallelizations of SSGD fail to scale due to huge communication overhead. During an SGD epoch, these methods send data proportional to one of the dimensions of the rating matrix. We propose a framework for scalable SSGD through significantly reducing the communication overhead via exchanging point-to-point messages utilizing the sparsity of the input matrix. We provide formulas to represent the essential communication for correctly performing parallel SSGD and we propose a dynamic programming algorithm for efficiently computing them to establish the point-to-point message schedules. This scheme, however, significantly increases the number of messages sent by a processor per epoch from O(K) to O(K^2) for a K-processor system which might limit the scalability. To remedy this, we propose a Hold-and-Combine strategy to limit the upper-bound on the number of messages sent per processor to O(KlgK). We also propose a hypergraph partitioning model that correctly encapsulates reducing the communication volume. Experimental results show that the framework successfully achieves a scalable distributed SSGD through significantly reducing the communication overhead. Our code is publicly available at: github.com/nfabubaker/CESSGD

The recent expansion of the Internet of things (IoT) and its applications continues to dramatically change the IT services supporting manufacturing, energy management, intelligent transportation, smart cities, etc. These IT services often interconnect Internet-based Cloud systems with IoT or intelligent smart devices held by organisations. As IoT devices are resource-limited, the security of their communication is provided by lightweight cryptography that could consequently lead to the devices being insecure against a quantum adversary. On the other hand, Internet Cloud systems are resource-rich and could easily support post-quantum cryptography. This paper presents a proxy re-encryption scheme that translates ciphertexts of a classical lightweight encryption algorithm into their post quantum equivalent and vice versa. More precisely, we propose a lattice-based ABE-IBE (L ABE-IBE) proxy re-encryption scheme that transforms a classical lightweight ciphertext into a lattice based ciphertext. We compare our proposed L ABE-IBE proxy re-encryption scheme with classical proxy re-encryption (ABEIBE) schemes. We discuss the security and efficiency of our construction.

While the technique of Deep Neural Networks (DNNs) has been instrumental in achieving state-of-the-art results for various Natural Language Processing (NLP) tasks, recent works have shown that the decisions made by DNNs cannot always be trusted. Recently Explainable Artificial Intelligence (XAI) methods have been proposed as a method for increasing DNN’s reliability and trustworthiness. These XAI methods are however open to attack and can be manipulated in both white-box (gradient-based) and black-box (perturbation-based) scenarios. Exploring novel techniques to attack and robustify these XAI methods is crucial to fully understand these vulnerabilities. In this work, we propose Tamp-X—a novel attack which tampers the activations of robust NLP classifiers forcing the state-of-the-art white-box and black-box XAI methods to generate misrepresented explanations. To the best of our knowledge, in current NLP literature, we are the first to attack both the white-box and the black-box XAI methods simultaneously. We quantify the reliability of explanations based on three different metrics—the descriptive accuracy, the cosine similarity, and the Lp norms of the explanation vectors. Through extensive experimentation, we show that the explanations generated for the tampered classifiers are not reliable, and significantly disagree with those generated for the untampered classifiers despite that the output decisions of tampered and untampered classifiers are almost always the same. Additionally, we study the adversarial robustness of the tampered NLP classifiers, and find out that the tampered classifiers which are harder to explain for the XAI methods, are also harder to attack by the adversarial attackers.

Recent works have highlighted how misinformation is plaguing our online social networks. Numerous algorithms on automated misinformation detection are centered around deep learning~(DL) which requires large data for training. However, privacy and ethical concerns reduce data sharing by stakeholders, impeding data-driven misinformation detection. Current data encryption techniques providing privacy guarantees cannot be naively extended to text inference with DL models, mainly due to the errors induced by stacked encrypted operations and polynomial approximations of the otherwise encryption-incompatible non-polynomial operations. In this paper, we show, formally and empirically, the effectiveness of (1) $L_2$ regularized training to reduce the overall error induced by approximate polynomial activations, and (2) sigmoid activation to regulate the error accumulated due to cascaded operations over encrypted data. We assume a federated learning-encrypted inference~(FL-EI) setup for text-based misinformation detection as a (secure and privacy-aware cloud) service, where classifiers are securely trained in FL framework and inference is performed on homomorphically encrypted data. We evaluate three architectures—Logistic Regression~(LR), Multilayer Perceptron~(MLP), and Self-Attention Network~(SAN)—on two public text-misinformation datasets with some interesting results, for example, by simply replacing ReLU activation with sigmoid, we were able to reduce the output error by $1750\times$ in the best case to $43.75\times$ in the worst case.

The possibility of creating vacuum robotics based on the polymer structures irradiated by an electron beam, in particular, polymer fibers, which provide high functional flexibility and a variety of states, is discussed. The possibility of using polymer fibers as different types of MEMS-like electromechanical elements is demonstrated - from elastic cantilevers to springs that change their state under the electron beam. Experimentally proved the presence of different functional types of fibers, correlating with their thickness, as well as the phenomenon of the fiber break. A number of exotic forms of dynamics have been demonstrated and a method for their detection has been developed using 2D Fourier spectra, integral spatial characteristics, time rsolved correlograms and wavelet transforms (visualized as the scaleograms / scalograms). Access barcodes for the full video records of the corresponding experiments are provided.

A current trend in the development of the applied measuring techniques is the design of integrated analytical systems on a chip, in particular, labs on a chip based on various physical principles of the substance analysis. Recently, the term “lab-on-a-chip” implied analytical microsystems, requiring macro-scopic readers for the data obtained, which strongly limited the possibilities of their introduction into the field studies and routine analytical measure-ments. However, a progress in the design of telemetric labs on a chip, in-cluding those capable of broadcasting the analytical signal to the receivers with positional sensitivity provided by the use of the labeled counting cham-bers, made it possible to use such devices beyond the scientific laboratories: in the field practice or in monitoring of the natural systems and artificial structures. This can be applied not only to the chemical and biological field measurements, but also to the mechanics, in particular, to geomechanics or soil mechanics. This work briefly considers the results obtained in the Rus-sian laboratory of the second author during the period up to 2011-2013, which have not been published earlier due to the know-how restrictions.

We propose a novel method to detect underground cavities using cross-borehole ground-penetrating radar (GPR).

A procedure to identify single amino acids (AAs) in the bulk or with a few molecules from a single binary measurement is introduced. Combined with Edman degradation (or other method) it can be used to sequence a peptide or identify the parent protein from a partial sequence. This binary/digital approach is centered on the superspecificity property of transfer RNAs (tRNAs), it differs markedly from conventional and recent single molecule (SM) sequencing methods based on analog measurements. Each of 20 (or more) terminal residues cleaved from 20 (or more) copies of a peptide enters a different cavity with a unique tRNA; tRNA charging (or binding with AA) occurs only in the cavity with the cognate AA. The bound AA or the AA separated from the tRNA is detected with a single binary measurement; its identity is known from the position of the single high bit in the resulting 20-bit output. Alternatively a 20-stage pipeline can be used with sparse samples. Detection of the bound AA can be done optically by tagging the AAs with a fluorescent dye, of the freed AA electrically with a nanopore. Necessary conditions for accurate AA identification are satisfied in principle; related computations and simulation results are presented. A modified version that can be used for de novo sequencing in parallel of large numbers of peptides immobilized on a glass slide with the tRNAs carrying a fluorescent tag is mooted. Both methods can be used for protein identification from partial sequences containing 2 or 3 AA types by using only the corresponding tRNAs. Potential implementation issues are discussed. With a straightforward extension post-translational modifications (PTMs) in single AAs can be detected. Such detection is horizontal: it is only necessary to search among the PTMs for a given AA, PTMs of the other 19 AAs are not involved.

Drug repurposing can accelerate the identification of effective compounds for clinical use against SARS-CoV-2, with the advantage of pre-existing clinical safety data and an established supply chain. RNA viruses such as SARS-CoV-2 manipulate cellular pathways and induce reorganization of subcellular structures to support their life cycle. These morphological changes can be quantified using bioimaging techniques. In this work, we developed DEEMD: a computational pipeline using deep neural network models within a multiple instance learning framework, to identify putative treatments effective against SARS-CoV-2 based on morphological analysis of the publicly available RxRx19a dataset. This dataset consists of fluorescence microscopy images of SARS-CoV-2 non-infected cells and infected cells, with and without drug treatment. DEEMD first extracts discriminative morphological features to generate cell morphological profiles from the non-infected and infected cells. These morphological profiles are then used in a statistical model to estimate the applied treatment efficacy on infected cells based on similarities to non-infected cells. DEEMD is capable of localizing infected cells via weak supervision without any expensive pixel-level annotations. DEEMD identifies known SARS-CoV-2 inhibitors, such as Remdesivir and Aloxistatin. supporting the validity of our approach. DEEMD can be explored for use on other emerging viruses and datasets to rapidly identify candidate antiviral treatments in the future. Our implementation is available online at https://github.com/Sadegh-Saberian/DEEMD.

Tensor computations are important mathematical operations for applications that rely on multidimensional data. The tensor–vector multiplication (TVM) is the most memory-bound tensor contraction in this class of operations. This paper proposes an open-source TVM algorithm which is much simpler and efficient than previous approaches, making it suitable for integration in the most popular BLAS libraries available today. Our algorithm has been written from scratch and features unit-stride memory accesses, cache awareness, mode obliviousness, full vectorization and multi-threading as well as NUMA awareness for non-hierarchically stored dense tensors. Numerical experiments are carried out on tensors up to order 10 and various compilers and hardware architectures equipped with traditional DDR and high bandwidth memory (HBM). For large tensors the average performance of the TVM ranges between 62\% and 76\% of the theoretical bandwidth for NUMA systems with DDR memory and remains independent of the contraction mode. On NUMA systems with HBM the TVM exhibits some mode dependency but manages to reach performance figures close to peak values. Finally, the higher-order power method is benchmarked with the proposed TVM kernel and delivers on average between 58\% and 69\% of the theoretical bandwidth for large tensors.