With the advent of industry 4.0, data-driven techniques have gained a lot of popularity for machine condition monitoring, ensuring reliable and safe operation of the machines. In most practical application scenarios, domain discrepancy may arise between the training (source domain) and test (target domain) data due to various factors like changes in the operating conditions, different sensor locations, etc. Traditional data-driven techniques fail to address this domain shift, and hence domain adaptation techniques are required to ensure reliable performance. This work presents an unsupervised domain adaptation method where labeled data is available only in the source domain via subspace interpolation using deep dictionary learning. Deep dictionaries learn rich representations from the data and hence are used for subspace interpolation to capture the domain shift and form a shared feature space for cross-domain analysis. The proposed method is evaluated for the challenging scenario of adaptation between different but related machines. Experimental results obtained with two publicly available bearing fault datasets are promising; the proposed method significantly outperforms all the state-of-the-art methods.

Quantum Image Representation is researched from last few years, and more active in the recent past. Set to examine how these representations would be useful for Image Processing in a quantum way, we considered the Quantum Machine Learning problem of image classification in this paper. Encouraging results have been provided on classifying benchmark datasets of grayscale and colour images using two different classifiers and their combination. Multiclass classification performance has also been tested.

Quantum and Classical computers continue to work together in tight cooperation to solve difficult problems. The combination is thus suggested in recent times for decoding the Low Density Parity Check (LDPC) codes, for the next generation Wireless Communication systems. In this paper we have worked out the Quadratic Unconstrained Binary Optimization (QUBO) formulation for Rayleigh Fading channels for two different scenarios- channel state fully known and not known. The resultant QUBO are solved using D-Wave 2000Q Quantum Annealer and the outputs from the Annealer are classically postprocessed, invoking the notion of diversity. Simple minimum distance decoding of the available copies of the outputs led to improved performance, compared to picking the minimum-energy solution in terms of Bit Error Rate (BER). Apart from providing these results and the comparisons to fully classical Simulated Annealing (SA) and the traditional Belief Propagation (BP) based strategies, some remarks about diversity due to quantum processing are also spelt out.

Enormous activity in the Quantum Computing area has resulted in it being considered, together with classical computers, for solving different difficult problems - including those of applied nature. An attempt is made in this work to assemble a pipeline consisting of both quantum and classical processing blocks for the task of image classification and segmentation, keeping in mind the present limitations of the gate-model quantum computers. It is based on the work done for the recent BMW Quantum Computing Challenge related to Automotive Industry. The pipeline handles the real-life sized images as the input and output, rather than the toy-sized examples prevalent in Quantum Computing literature. Apart from breaking down the problem to modules, some of which can be accommodated in the existing simulators and hardware, simplifications of the relevant quantum algorithms are also carried out. Its functionality and utility are brought out by applying it to surface crack segmentation on the popular Kaggle Surface Crack Detection data set. The results of the paper are not only limited to simulations, but also involve running models on Noisy, Intermediate-Scale Quantum processors through AWS. In its entirety, this work may lay the groundwork for quantum/quantum-enhanced image segmentation, and providing interested researchers with a stepping-stone in that direction, as the results demonstrate the efficacy of the proposed method, even with simple versions of the quantum modules.

Fault diagnosis plays a vital role in ensuring the normal operation of the machine and safe production. In recent years, data-driven techniques have gained a lot of popularity for machine fault diagnosis. But most of these techniques assume the training and test data have the same distribution. However, in most practical application scenarios, domain discrepancy can be observed between the training (source) and test (target) data due to different factors like changes in the operating conditions, different sensor locations, etc. Classical approaches fail to address such domain discrepancy, which leads to poor performance. The problem becomes more challenging when the target is completely unlabeled. To address this scenario, domain adaptation techniques are used to transfer the knowledge learned from the labeled source domain to the unlabeled target domain. Recently, adversarial network based domain adaptation has been extensively explored for fault diagnosis. But the adversarial loss alone does not guarantee the translation of the source to the desired target domain (class consistent). Here, we propose to use cycle-consistency loss employing 1D-CycleGAN for learning the source to target mapping for unsupervised adaptation for bearing fault diagnosis. The proposed method is evaluated for two different scenarios, with the source and target from (i) same machine but different working conditions and (ii) different but related machines. Experimental results show that while the proposed method performs comparable to the best-performing benchmark for the first case, it significantly outperforms all the state-of-the-art methods for the challenging second case.

Insights from multiple imaging modalities have recently been applied in solving many computer vision related applications. In this paper, we model the cross-modal dependencies between different modalities for Multimodal Image Super-Resolution (MISR), i.e., enhance the Low Resolution (LR) image of target modality with the guidance of a High Resolution (HR) image from another modality. We introduce a joint optimization based transform learning frame-work referred to as Joint Coupled Transform Learning (JCTL) to combine the information from multiple modalities to generate the HR image of the target modality. All the necessary intermediate steps and the corresponding closed form solution updates are pro-vided. The performance of the proposed JCTL is benchmarked against the state-of-the-art MISR approaches on different multi-modal datasets with different upscaling factors. The results show better performance with the proposed JCTL approach compared to other state-of-the-art techniques both in terms of PSNR and SSIM.

The performance of existing methods for multisensor fusion are severely affected by the lack of significant amount of labeled data. In most practical scenarios, the amount of unlabeled data is huge in comparison to labeled data. To address this problem, a novel autoencoder based multi-sensor fusion framework for semi-supervised learning is proposed in this work. Here, both labeled and unlabeled data are used for learning the latent representation from each sensor. Subsequently, the latent representation of all the sensors are combined to perform classification. A joint optimization formulation is presented for learning the sensor-specific latent representation, their encoder and decoder weights and the classification weights together. This ensures discriminative features to be learnt from individual sensors that aids in classification. The requisite solution steps and the closed form updates for the joint learning of all the parameters are given. Experiment results presented on two datasets from different domains demonstrate the generalizability and superior performance of the proposed AutoFuse compared to state-of-the-art methods with relatively less complexity and the ability to work with partially annotated data.

Generative Models have always attracted the attention of Machine Learning research community; they are useful and also generally harder than their discriminative counterparts. In these models, we would be looking into learning the probability distribution of the input and sampling from that to generate new data samples. Since quantum computing and algorithms are inherently random, they can facilitate a natural framework in this situation. But, getting a suitable gate circuit to achieve the requisite quantum state which by repeated preparation and measurement leads to the sought-after data samples is not trivial. In this paper, we propose a quantum circuit which has a flavor of Variational Autoencoder with the usual visible and hidden nodes for input data and latent distribution. The encoder portion comprises of a suitably chosen parameterized phase ansatz and Inverse Quantum Fourier Transform blocks. Depending on whether the measurement is carried out on the hidden nodes or not, the decoder circuit, which is just not the inverse of the encoder in our case, is configured. The Kullback-Leibler Divergence is used train the circuit towards the required input distribution. Numerical results presented demonstrate the correct functionality of the approach.

Load Disaggregation has gained much popularity in the recent times, owing to the advantages it brings to energy utility companies. Many modeling techniques ranging from Dictionary Learning to HMM-based techniques to Neural Network based modeling have been proposed in the literature to solve this problem. However, scalability and computational lightness, have been two main areas of concern associated with the problem modeling. In this work, the authors propose to use Multi-Label Auto-Encoder architecture to solve this problem. The proposed architecture incurs minimum instrumentation cost, which makes it truly non-intrusive. The use of superposed appliance class labels of interest in the discriminative penalizing term of the architecture, ensures that disaggregation is achieved without the need to train a separate model for each appliance class of interest.

In this paper, we present a kernelized dictionary learning framework for carrying out regression to model signals having a complex nonlinear nature. A joint optimization is carried out where the regression weights are learnt together with the dictionary and coefficients. Relevant formulation and dictionary building steps are provided. To demonstrate the effectiveness of the proposed technique, elaborate experimental results using different real-life datasets are presented. The results show that non-linear dictionary is more accurate for data modeling and provides significant improvement in estimation accuracy over the other popular traditional techniques especially when the data is highly non-linear.

Present day industrial machines are equipped with many internal sensors- both physical and virtual. Gleaning the information from these to ascertain the state of a process or a subsystem as well as to predict any relevant failures in the manufacturing is all the more important in the envisaged fourth industrial revolution. Getting the robust estimation of the requisite quantities as a function of the available measurements is a highly challenging task due to the interplay of many factors. The problem is further aggravated when the data is both limited and poorly annotated. Signal processing can play a vital role in these scenarios to arrive at useful features for the robust estimation and prediction. Focusing on the prediction of tool wear in Computer Numerical Control (CNC) machines, the paper proposes deriving novel features, including those based on Singular Spectrum Analysis and Graph Total Variation. By considering the appropriate combination of these features as arguments to the requisite underlying function, the latter is learned through traditional machine learning structures. The usefulness of the suggested overall technique based on the said blend is brought out by providing the results obtained on the real-life data.

The deployment of smart meters by utilities holds the promise of improvements in operational efficiency, reliability and cost savings. With power measurements from smart meters, utilities can deploy innovative programs that allow end users to better control their energy usage while simultaneously reducing peak demand across the grid. In this paper, to develop data analysis tools for applications enabling monitoring and control of energy, a systems approach is taken, comprising of modeling, measurement, calibration and inference on the energy data collected from end users. A combination of analysis and synthesis for deriving data and measurement models calibrated to the aggregate power under measurement allows detection and estimation of features of individual appliances. Test results on disaggregation of power waveforms using the publicly available REDD data sets show promising results. The generic modeling and optimization framework can be used in the design and deployment of cyber physical energy systems for monitoring and control of energy resources.

The ubiquitous location sensing trend has increased the demand for low-cost GPS receivers; but their energy needs are still too high. For delay-tolerant applications, investigations show that significant energy saving can be obtained by offloading a few milliseconds of raw signal samples and leveraging the greater processing power of the cloud for obtaining a position fix. In an attempt to reduce the energy cost of this data offloading operation, we propose SparseGPS+. Based on the sparse decomposition model, it overcomes many limitations of SparseGPS [1] to yield better signal-to-noise ratio and detection accuracy; which translates to 30% more energy savings compared to the state-of-the-art.

The sparse error correction is intimately related to Compressive Sensing. Exploiting this connection, the paper proposes an error correction scheme pivoted on partial Fourier matrix constructed using cyclic difference sets. The scheme is elegant in terms of computationally efficient encoding and decoding as well as guaranteed error correction.

Ubiquitous integration of high-speed WLANs with wide-range 3GPP systems results in the service extension of the backbone cellular network. This paper envisions such heterogeneous wireless network architecture by integrating IEEE 802.11p VANETs with 3GPP LTE to achieve seamless data connectivity for uninterrupted multimedia sessions amongst spatially-apart vehicular clusters. Issues on cluster head-based multicasting and QoS are explored in this paper. An adaptive multi-metric Cluster Head (CH) election mechanism is proposed to manage the VANET sub-clusters. In addition to this, construction of a 2-hop virtual overlay mesh-based shared multicast tree for lower-level multicasting within VANETs is discussed. Following this, the process of VANET-LTE upper-level communication is detailed, addressing the issues of CH and gateway handover, and resource allocation of the LTE eNB. The envisioned architecture enables the LTE to effectively schedule multimedia sessions based on the service requirements of the VANET gateways, thus satisfying QoS. Requisite simulation results are presented to evaluate the integrated network.

Complex Event Processing (CEP) has received wider acceptability due to its systematic and multilevel architecture driven concept approach. CEP is an emerging technology in the field of data processing and identifying patterns of interest from multiple streams of events. High levels of integrated self learning applications can be developed. CEP is used in development of applications which have to deal with voluminous streams of incoming data with the task of finding meaningful events or patterns of events, and respond to the events of interest in real time. In this paper a CEP based application for object detection tracking in a Wireless Sensor Network (WSN) environment is proposed. Also the detection of an intruder using semantic query processing is proposed. ESPER, an open source Complex Event Processing engine is used to develop the application.

In this paper, an efficient method for face recognition based on the Discrete Cosine Transform (DCT), Fisher Linear Discriminant (FLD) and classifier is presented. First, the dimensionality of the original face image is reduced using the DCT and illumination variations are alleviated by discarding the first few low-frequency DCT coefficients. FLD is applied to the selected DCT coefficients to discriminate the invariant facial features. The KNN classifier is used for the recognition of the faces using the features extracted from the FLD. Simulation results show that the proposed system achieves better performance with high training and high recognition rate as well as very good illumination robustness.