Search

Recent years have witnessed the success of deep neural networks in dealing with a plenty of practical problems. Dropout has played an essential role in many successful deep neural networks, by inducing regularization in the model training. In this paper, we present a new regularized training approach: Shakeout. Instead of randomly discarding units as Dropout does at the training stage, Shakeout randomly chooses to enhance or reverse each unit's contribution to the next layer. This minor modification of Dropout has the statistical trait: the regularizer induced by Shakeout adaptively combines L0 , L1 and L2 regularization terms. Our classification experiments with representative deep architectures on image datasets MNIST, CIFAR-10 and ImageNet show that Shakeout deals with over-fitting effectively and outperforms Dropout. We empirically demonstrate that Shakeout leads to sparser weights under both unsupervised and supervised settings. Shakeout also leads to the grouping effect of the input units in a layer. Considering the weights in reflecting the importance of connections, Shakeout is superior to Dropout, which is valuable for the deep model compression. Moreover, we demonstrate that Shakeout can effectively reduce the instability of the training process of the deep architecture. [ABSTRACT FROM AUTHOR]

Street scene understanding is an essential task for autonomous driving. One important step toward this direction is scene labeling, which annotates each pixel in the images with a correct class label. Although many approaches have been developed, there are still some weak points. First, many methods are based on the hand-crafted features whose image representation ability is limited. Second, they cannot label foreground objects accurately due to the data set bias. Third, in the refinement stage, the traditional Markov random filed inference is prone to over smoothness. For improving the above problems, this paper proposes a joint method of priori convolutional neural networks at superpixel level (called as “priori s-CNNs”) and soft restricted context transfer. Our contributions are threefold: 1) a priori s-CNNs model that learns priori location information at superpixel level is proposed to describe various objects discriminatingly; 2) a hierarchical data augmentation method is presented to alleviate data set bias in the priori s-CNNs training stage, which improves foreground objects labeling significantly; and 3) a soft restricted MRF energy function is defined to improve the priori s-CNNs model’s labeling performance and reduce the over smoothness at the same time. The proposed approach is verified on CamVid data set (11 classes) and SIFT Flow Street data set (16 classes) and achieves a competitive performance. [ABSTRACT FROM PUBLISHER]

Existing methodologies to count vehicles from a road image have depended upon both hand-crafted feature engineering and rule-based algorithms. These require many predefined thresholds to detect and track vehicles. This paper provides a supervised learning methodology that requires no such feature engineering. A deep convolutional neural network was devised to count the number of vehicles on a road segment based solely on video images. The present methodology does not regard an individual vehicle as an object to be detected separately; rather, it collectively counts the number of vehicles as a human would. The test results show that the proposed methodology outperforms existing schemes. [ABSTRACT FROM PUBLISHER]

Given a set of images that contain objects from a common category, object co-segmentation aims at automatically discovering and segmenting such common objects from each image. During the past few years, object co-segmentation has received great attention in the computer vision community. However, the existing approaches are usually designed with misleading assumptions, unscalable priors, or subjective computational models, which do not have sufficient robustness for dealing with complex and unconstrained real-world image contents. This paper proposes a novel two-stage co-segmentation framework, mainly for addressing the robustness issue. In the proposed framework, we first introduce the concept of union background and use it to improve the robustness for suppressing the image backgrounds contained by the given image groups. Then, we also weaken the requirement for the strong prior knowledge by using the background prior instead. This can improve the robustness when scaling up for the unconstrained image contents. Based on the weak background prior, we propose a novel MR-SGS model, i.e., manifold ranking with the self-learned graph structure, which can infer suitable graph structures in a data-driven manner rather than building the fixed graph structure relying on the subjective design. Such capacity is critical for further improving the robustness in inferring the foreground/background probability of each image pixel. Comprehensive experiments and comparisons with other state-of-the-art approaches can demonstrate the effectiveness of the proposed work. [ABSTRACT FROM PUBLISHER]

This paper describes the characterization of reflectarray unit cells using support vector machines (SVMs) to obtain fast and accurately the full matrix of reflection coefficients, which is used for an analysis of dual-polarized reflectarrays, demonstrating the performance of the model. First, a surrogate model of the reflectarray unit cell is obtained using SVMs. To this end, a set of random samples of the reflection coefficient matrix with a full-wave method of moments based on local periodicity (MoM-LP) is used to train the SVMs. To efficiently obtain the surrogate model, a novel strategy to accelerate the training process is presented, remarkably reducing computing time. Next, the model is tested against a different set of samples, obtaining an excellent agreement between the SVM model and MoM-LP simulations for all reflection coefficients, including the cross-coefficients. The surrogate model is then used for an efficient analysis of three reflectarrays with pencil beam for point-to-point communications, isoflux pattern for global Earth coverage, and a shaped beam for local multipoint distribution service application, showing excellent agreement in both copolar and crosspolar patterns between the SVM and MoM-LP simulations. Finally, the analysis is accelerated by a factor larger than three orders of magnitude using SVMs instead of MoM-LP. [ABSTRACT FROM PUBLISHER]

This paper presents a novel convolutional neural network (CNN) -based method for high-accuracy real-time car license plate detection. Many contemporary methods for car license plate detection are reasonably effective under the specific conditions or strong assumptions only. However, they exhibit poor performance when the assessed car license plate images have a degree of rotation, as a result of manual capture by traffic police or deviation of the camera. Therefore, we propose the a CNN-based MD-YOLO framework for multi-directional car license plate detection. Using accurate rotation angle prediction and a fast intersection-over-union evaluation strategy, our proposed method can elegantly manage rotational problems in real-time scenarios. A series of experiments have been carried out to establish that the proposed method outperforms over other existing state-of-the-art methods in terms of better accuracy and lower computational cost. [ABSTRACT FROM PUBLISHER]

This paper proposes a deep learning model to efficiently detect salient regions in videos. It addresses two important issues: 1) deep video saliency model training with the absence of sufficiently large and pixel-wise annotated video data and 2) fast video saliency training and detection. The proposed deep video saliency network consists of two modules, for capturing the spatial and temporal saliency information, respectively. The dynamic saliency model, explicitly incorporating saliency estimates from the static saliency model, directly produces spatiotemporal saliency inference without time-consuming optical flow computation. We further propose a novel data augmentation technique that simulates video training data from existing annotated image data sets, which enables our network to learn diverse saliency information and prevents overfitting with the limited number of training videos. Leveraging our synthetic video data (150K video sequences) and real videos, our deep video saliency model successfully learns both spatial and temporal saliency cues, thus producing accurate spatiotemporal saliency estimate. We advance the state-of-the-art on the densely annotated video segmentation data set (MAE of .06) and the Freiburg-Berkeley Motion Segmentation data set (MAE of .07), and do so with much improved speed (2 fps with all steps). [ABSTRACT FROM PUBLISHER]

Due to the advanced capabilities of the Internet of Vehicles (IoV) components such as vehicles, Roadside Units (RSUs) and smart devices as well as the increasing amount of data generated, Federated Learning (FL) becomes a promising tool given that it enables privacy-preserving machine learning that can be implemented in the IoV. However, the performance of the FL suffers from the failure of communication links and missing nodes, especially when continuous exchanges of model parameters are required. Therefore, we propose the use of Unmanned Aerial Vehicles (UAVs) as wireless relays to facilitate the communications between the IoV components and the FL server and thus improving the accuracy of the FL. However, a single UAV may not have sufficient resources to provide services for all iterations of the FL process. In this paper, we present a joint auction-coalition formation framework to solve the allocation of UAV coalitions to groups of IoV components. Specifically, the coalition formation game is formulated to maximize the sum of individual profits of the UAVs. The joint auction-coalition formation algorithm is proposed to achieve a stable partition of UAV coalitions in which an auction scheme is applied to solve the allocation of UAV coalitions. The auction scheme is designed to take into account the preferences of IoV components over heterogeneous UAVs. The simulation results show that the grand coalition, where all UAVs join a single coalition, is not always stable due to the profit-maximizing behavior of the UAVs. In addition, we show that as the cooperation cost of the UAVs increases, the UAVs prefer to support the IoV components independently and not to form any coalition. [ABSTRACT FROM AUTHOR]

Typical handheld controllers for interaction in virtual reality (VR) have fixed shapes and sizes, regardless of what visual objects they represent. Resolving this crossmodal incongruence with a shape-changing interface is our long-term goal. In this paper, we seek to find a length perception model that considers the moment of inertia (MOI) and diameter of a handheld object based on the concept of dynamic touch. Such models serve as a basis for computational algorithms for shape changing. We carried out two perceptual experiments. In Experiment 1, we measured the perceived lengths of 24 physical objects with different MOIs and diameters. Then we obtained a length perception model to reproduce the desired perceived length with a handheld controller. In Experiment 2, we validated our model in a crossmodal matching scenario, where a visual rod was matched to a haptic rod in terms of the perceived length. Our results contribute to understanding the relationship between the perceived length and physical properties of a handheld object and designing shape-changing algorithms to render equivalent visual and haptic sensory cues for length perception in VR. [ABSTRACT FROM AUTHOR]

In comparison to single-model based trackers, the model-ensembled tracking strategy has shown a substantial adaptivity in handling various tracking challenges. As the performance of the tracker has been improved by combining different model outputs linearly, the insufficient consideration of each ensemble member’s contribution still limits the tracking performance to be further enhanced. As the performance of the tracker has been improved by combining different model outputs linearly, the insufficient consideration of each ensemble member’s contribution still limits the tracking performance to be further enhanced. In this paper, a tracking strategy based on multiple instance models regression (MIMRT) is proposed with a unified ensembling scheme. By formulating the tracking initialization with an instance model, the encoding process for an object’s specific detail is performed corresponding to the samples in each frame. The advantage of this operation is to guarantee the model frame-wise discrimination of short-term training, as well as to evaluate the reliability of each instance model by utilizing the long-lifetime samples obtained throughout the whole tracking procedure. To finalize the proposed tracking, all the independent instance models attached to the learned regression coefficients are ensembled with respect to the long-lifetime samples. This also effectively bridges the instance model as a latent variable to investigate a semantic association between the tracking model and the overall samples. A comprehensive experiment has shown that the proposed tracker is able to achieve superior performances compared to the state-of-art tracking approaches on both short-term datasets (e.g., OTB2013, OTB100, VOT2016, UAV123) and long-term dataset (UAV20L). [ABSTRACT FROM AUTHOR]

In this paper, the problem of energy efficient transmission and computation resource allocation for federated learning (FL) over wireless communication networks is investigated. In the considered model, each user exploits limited local computational resources to train a local FL model with its collected data and, then, sends the trained FL model to a base station (BS) which aggregates the local FL model and broadcasts it back to all of the users. Since FL involves an exchange of a learning model between users and the BS, both computation and communication latencies are determined by the learning accuracy level. Meanwhile, due to the limited energy budget of the wireless users, both local computation energy and transmission energy must be considered during the FL process. This joint learning and communication problem is formulated as an optimization problem whose goal is to minimize the total energy consumption of the system under a latency constraint. To solve this problem, an iterative algorithm is proposed where, at every step, closed-form solutions for time allocation, bandwidth allocation, power control, computation frequency, and learning accuracy are derived. Since the iterative algorithm requires an initial feasible solution, we construct the completion time minimization problem and a bisection-based algorithm is proposed to obtain the optimal solution, which is a feasible solution to the original energy minimization problem. Numerical results show that the proposed algorithms can reduce up to 59.5% energy consumption compared to the conventional FL method. [ABSTRACT FROM AUTHOR]

Deep learning models are being deployed in many mobile intelligent applications. End-side services, such as intelligent personal assistants, autonomous cars, and smart home services often employ either simple local models on the mobile or complex remote models on the cloud. However, recent studies have shown that partitioning the DNN computations between the mobile and cloud can increase the latency and energy efficiencies. In this paper, we propose an efficient, adaptive, and practical engine, JointDNN, for collaborative computation between a mobile device and cloud for DNNs in both inference and training phase. JointDNN not only provides an energy and performance efficient method of querying DNNs for the mobile side but also benefits the cloud server by reducing the amount of its workload and communications compared to the cloud-only approach. Given the DNN architecture, we investigate the efficiency of processing some layers on the mobile device and some layers on the cloud server. We provide optimization formulations at layer granularity for forward- and backward-propagations in DNNs, which can adapt to mobile battery limitations and cloud server load constraints and quality of service. JointDNN achieves up to 18 and 32 times reductions on the latency and mobile energy consumption of querying DNNs compared to the status-quo approaches, respectively. [ABSTRACT FROM AUTHOR]

Given a spatial raster framework with explanatory feature layers, a spatial contextual layer (e.g., a potential field), as well as a set of training samples with class labels, the spatial prediction problem aims to learn a model that can predict a class layer. The problem is important in societal applications such as flood extent mapping for disaster response and national water forecasting, but is challenging due to the noise, obstacles, and heterogeneity in feature maps, implicit spatial dependency between locations based on the contextual layer (e.g., gradient directions on a potential field), and the large number of sample locations. Existing work often assumes undirected spatial dependency, or directed dependency with a total order, and thus cannot reflect complex directed dependency with a partial order. In contrast, we recently proposed geographical hidden Markov tree, a probabilistic graphical model that generalizes the common hidden Markov model from a one-dimensional sequence to a two-dimensional map. Partial order class dependency is incorporated in the hidden class layer with a reverse tree structure. We also investigated computational algorithms for reverse tree construction, model parameter learning and class inference. This paper extends our recent model with overlaying class nodes between observation nodes and underlying hidden class nodes. The additional overlaying class layer makes the model more robust to large scale feature obstacles. We also proposed corresponding learning and inference methods. Extensive evaluations on real world datasets show that our models outperform multiple baselines in flood mapping applications, our algorithms are scalable on large data sizes, and the proposed extension enhances classification performance. [ABSTRACT FROM AUTHOR]

The goodness of the features extracted from the instances and the number of training instances are two key components in machine learning, and building an effective model is largely affected by these two factors. Acquiring a large number of training instances is very expensive in some situations such as in the medical domain. Designing a good feature set, on the other hand, is very hard and often requires domain expertise. In computer vision, image descriptors have emerged to automate feature detection and extraction; however, domain-expert intervention is typically needed to develop these descriptors. The aim of this paper is to utilize genetic programming to automatically construct a rotation-invariant image descriptor by synthesizing a set of formulas using simple arithmetic operators and first-order statistics, and determining the length of the feature vector simultaneously using only two instances per class. Using seven texture classification image datasets, the performance of the proposed method is evaluated and compared against eight domain-expert hand-crafted image descriptors. Quantitatively, the proposed method has significantly outperformed, or achieved comparable performance to, the competitor methods. Qualitatively, the analysis shows that the descriptors evolved by the proposed method can be interpreted. [ABSTRACT FROM PUBLISHER]

Most of existing models for facial behavior analysis rely on generic classifiers, which fail to generalize well to previously unseen data. This is because of inherent differences in source (training) and target (test) data, mainly caused by variation in subjects’ facial morphology, camera views, and so on. All of these account for different contexts in which target and source data are recorded, and thus, may adversely affect the performance of the models learned solely from source data. In this paper, we exploit the notion of domain adaptation and propose a data efficient approach to adapt already learned classifiers to new unseen contexts. Specifically, we build upon the probabilistic framework of Gaussian processes (GPs), and introduce domain-specific GP experts (e.g., for each subject). The model adaptation is facilitated in a probabilistic fashion, by conditioning the target expert on the predictions from multiple source experts. We further exploit the predictive variance of each expert to define an optimal weighting during inference. We evaluate the proposed model on three publicly available data sets for multi-class (MultiPIE) and multi-label (DISFA, FERA2015) facial expression analysis by performing adaptation of two contextual factors: “where” (view) and “who” (subject). In our experiments, the proposed approach consistently outperforms: 1) both source and target classifiers, while using a small number of target examples during the adaptation and 2) related state-of-the-art approaches for supervised domain adaptation. [ABSTRACT FROM PUBLISHER]

In this paper, an efficient offline signature verification method based on an interval symbolic representation and a fuzzy similarity measure is proposed. In the feature extraction step, a set of local binary pattern-based features is computed from both the signature image and its under-sampled bitmap. Interval-valued symbolic data is then created for each feature in every signature class. As a result, a signature model composed of a set of interval values (corresponding to the number of features) is obtained for each individual’s handwritten signature class. A novel fuzzy similarity measure is further proposed to compute the similarity between a test sample signature and the corresponding interval-valued symbolic model for the verification of the test sample. To evaluate the proposed verification approach, a benchmark offline English signature data set (GPDS-300) and a large data set (BHSig260) composed of Bangla and Hindi offline signatures were used. A comparison of our results with some recent signature verification methods available in the literature was provided in terms of average error rate and we noted that the proposed method always outperforms when the number of training samples is eight or more. [ABSTRACT FROM PUBLISHER]

Space mapping (SM) is a recognized method for speeding up electromagnetic (EM) optimization. Existing SM approaches are mostly based on sequential computation mechanism. This paper proposes a parallel SM method for EM optimization. In the proposed method, the surrogate model developed in each iteration is trained to match the fine model at multiple points simultaneously. Multi-point training and SM enables the surrogate model to be valid in a larger neighborhood than that in standard SM. The proposed formulation of multi-point surrogate model training is inherently suited to and implemented through parallel computation. This includes multiple fine model evaluation in parallel and multi-point surrogate training using a parallel algorithm. Our proposed method further reduces the number of SM iterations and speeds up the optimization process in comparison with the standard SM. This technique is illustrated by three microwave filter examples. [ABSTRACT FROM PUBLISHER]

Optimization techniques have been used for many years in the formulation and solution of computational problems arising in speech and language processing. Such techniques are found in the Baum-Welch, extended Baum-Welch (EBW), Rprop, and GIS algorithms, for example. Additionally, the use of regularization terms has been seen in other applications of sparse optimization. This paper outlines a range of problems in which optimization formulations and algorithms play a role, giving some additional details on certain application problems in machine translation, speaker/language recognition, and automatic speech recognition. Several approaches developed in the speech and language processing communities are described in a way that makes them more recognizable as optimization procedures. Our survey is not exhaustive and is complemented by other papers in this volume. [ABSTRACT FROM PUBLISHER]

In federated learning (FL), devices contribute to the global training by uploading their local model updates via wireless channels. Due to limited computation and communication resources, device scheduling is crucial to the convergence rate of FL. In this paper, we propose a joint device scheduling and resource allocation policy to maximize the model accuracy within a given total training time budget for latency constrained wireless FL. A lower bound on the reciprocal of the training performance loss, in terms of the number of training rounds and the number of scheduled devices per round, is derived. Based on the bound, the accuracy maximization problem is solved by decoupling it into two sub-problems. First, given the scheduled devices, the optimal bandwidth allocation suggests allocating more bandwidth to the devices with worse channel conditions or weaker computation capabilities. Then, a greedy device scheduling algorithm is introduced, which selects the device consuming the least updating time obtained by the optimal bandwidth allocation in each step, until the lower bound begins to increase, meaning that scheduling more devices will degrade the model accuracy. Experiments show that the proposed policy outperforms state-of-the-art scheduling policies under extensive settings of data distributions and cell radius. [ABSTRACT FROM AUTHOR]

As one of the indispensable components in convolutional neural network (CNN), loss function assists in updating parameters of CNN models during the training phase. Generally, different loss functions can assist convolutional neural network (CNN) to learn different feature representations, and different feature representations can be treated as different knowledge learned from objects. In this paper we introduce a novel training framework, namely Loss Transferring (LT), to improve the generalization ability of CNN. LT contains multiple training phases, and each training phase uses a different loss function. Under this framework, CNN models can combine different knowledge of objects by transferring the knowledge learned via one loss function to another. LT contains two components, i.e., loss function set and training strategy. In order to build appropriate loss function set, we establish two basic guides. And according to these basic guides, we design a new loss function in the last layer of CNN models (layer before softmax operation), namely Near Classifier Hyper-Plane (N-CHP) loss, which makes the learned object features belonging to the same category have the minimum intra-class distance and be near the classifier hyper-plane. Based on the two loss function set $\{\text {MSE, softmax}\}$ and $\{\text {N-CHP, softmax}\}$ , we setup two specific training methods, $LT_{\text {MSE, softmax}}$ and $LT_{\text {N-CHP, softmax}}$ , which can be universally applied to different CNN models with low additional computation cost. Meanwhile, two training strategies, multi-phase strategy 1 and multi-phase strategy 2, are further proposed to improve the training efficiency of LT. Extensive experimental results on shallow, moderate and deep models with four benchmark datasets, including MNIST, SVHN, CIFAR-10 and CIFAR-100, demonstrate that CNN models can bring obvious performance improvements when working with $LT_{\text {MSE, softmax}}$ and $LT_{\text {N-CHP, softmax}}$ , which verifies the effectiveness of LT and the proposed two basic guides. [ABSTRACT FROM AUTHOR]

Although deep features have achieved the state-of-the-art performance in action recognition recently, the hand-crafted shallow features still play a critical role in characterizing human actions for taking advantage of visual contents in an intuitive way such as edge features. Therefore, the shallow features can serve as auxiliary visual cues supplementary to deep representations. In this paper, we propose a discriminative subspace learning model (DSLM) to explore the complementary properties between the hand-crafted shallow feature representations and the deep features. As for the RGB action recognition, this is the first work attempting to mine multi-level feature complementaries by the multi-view subspace learning scheme. To sufficiently capture the complementary information among heterogeneous features, we construct the DSLM by integrating the multi-view reconstruction error and classification error into an unified objective function. To be specific, we first use Fisher Vector to encode improved dense trajectories (iDT+FV) for shallow representations and two-stream convolutional neural network models (T-CNN) for generating deep features. Moreover, the presented DSLM algorithm projects multi-level features onto a shared discriminative subspace with the complementary information and discriminating capacity simultaneously incorporated. Finally, the action types of test samples are identified by the margins from the learned compact representations to the decision boundary. The experimental results on three datasets demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

The direct analysis of 3D Optical Coherence Tomography (OCT) volumes enables deep learning models (DL) to learn spatial structural information and discover new bio-markers that are relevant to glaucoma. Downsampling 3D input volumes is the state-of-art solution to accommodate for the limited number of training volumes as well as the available computing resources. However, this limits the network's ability to learn from small retinal structures in OCT volumes. In this paper, our goal is to improve the performance by providing guidance to DL model during training in order to learn from finer ocular structures in 3D OCT volumes. Therefore, we propose an end-to-end attention guided 3D DL model for glaucoma detection and estimating visual function from retinal structures. The model consists of three pathways with the same network architecture but different inputs. One input is the original 3D-OCT cube and the other two are computed during training guided by the 3D gradient class activation heatmaps. Each pathway outputs the class-label and the whole model is trained concurrently to minimize the sum of losses from three pathways. The final output is obtained by fusing the predictions of the three pathways. Also, to explore the robustness and generalizability of the proposed model, we apply the model on a classification task for glaucoma detection as well as a regression task to estimate visual field index (VFI) (a value between 0 and 100). A 5-fold cross-validation with a total of 3782 and 10,370 OCT scans is used to train and evaluate the classification and regression models, respectively. The glaucoma detection model achieved an area under the curve (AUC) of 93.8% compared with 86.8% for a baseline model without the attention-guided component. The model also outperformed six different feature based machine learning approaches that use scanner computed measurements for training. Further, we also assessed the contribution of different retinal layers that are relevant to glaucoma. The VFI estimation model achieved a Pearson correlation and median absolute error of 0.75 and 3.6%, respectively, for a test set of size 3100 cubes. [ABSTRACT FROM AUTHOR]

To leverage data and computation capabilities of mobile devices, machine learning algorithms are deployed at the network edge for training artificial intelligence (AI) models, resulting in the new paradigm of edge learning. In this paper, we consider the framework of partitioned edge learning for iteratively training a large-scale model using many resource-constrained devices (called workers). To this end, in each iteration, the model is dynamically partitioned into parametric blocks, which are downloaded to worker groups for updating using data subsets. Then, the local updates are uploaded to and cascaded by the server for updating a global model. To reduce resource usage by minimizing the total learning-and-communication latency, this work focuses on the novel joint design of parameter (computation load) allocation and bandwidth allocation (for downloading and uploading). Two design approaches are adopted. First, a practical sequential approach, called partially integrated parameter-and-bandwidth allocation (PABA), yields two schemes, namely bandwidth aware parameter allocation and parameter aware bandwidth allocation. The former minimizes the load for the slowest (in computing) of worker groups, each training a same parametric block. The latter allocates the largest bandwidth to the worker being the latency bottleneck. Second, PABA are jointly optimized. Despite it being a nonconvex problem, an efficient and optimal solution algorithm is derived by intelligently nesting a bisection search and solving a convex problem. Experimental results using real data demonstrate that integrating PABA can substantially improve the performance of partitioned edge learning in terms of latency (by e.g., 46%) and accuracy (by e.g., 4% given the latency of 100 seconds). [ABSTRACT FROM AUTHOR]

This paper presents a new semi-supervised probabilistic density-based regression approach, called Semi-supervised Weighted Gaussian Regression (SWGR), for the soft sensing of nonlinear and multimode industrial processes given a limited number of labeled data samples. In SWGR, different weights are assigned to each training sample based on their similarities to a query sample. Then a local weighted Gaussian density is built for capturing the joint probability of historical samples around the query sample. The training process of parameters in SWGR incorporates both labeled and unlabeled data samples via a maximum likelihood estimation algorithm. In this way, the soft sensor model is able to approximate the nonlinear mechanics of input and output variables and remedy the insufficiency of labeled samples. At last, the output prediction as well as the uncertainty of prediction can be obtained by the conditional distribution. Two case studies validate that the proposed semi-supervised soft sensing method outperforms some recent methods. [ABSTRACT FROM AUTHOR]

While machine-type communication (MTC) devices generate considerable amounts of data, they often cannot process the data due to limited energy and computational power. To empower MTC with intelligence, edge machine learning has been proposed. However, power allocation in this paradigm requires maximizing the learning performance instead of the communication throughput, for which the celebrated water-filling and max-min fairness algorithms become inefficient. To this end, this paper proposes learning centric power allocation (LCPA), which provides a new perspective on radio resource allocation in learning driven scenarios. By employing 1) an empirical classification error model that is supported by learning theory and 2) an uncertainty sampling method that accounts for different distributions at users, LCPA is formulated as a nonconvex nonsmooth optimization problem, and is solved using a majorization minimization (MM) framework. To get deeper insights into LCPA, asymptotic analysis shows that the transmit powers are inversely proportional to the channel gains, and scale exponentially with the learning parameters. This is in contrast to traditional power allocations where quality of wireless channels is the only consideration. Last but not least, a large-scale optimization algorithm termed mirror-prox LCPA is further proposed to enable LCPA in large-scale settings. Extensive numerical results demonstrate that the proposed LCPA algorithms outperform traditional power allocation algorithms, and the large-scale optimization algorithm reduces the computation time by orders of magnitude compared with MM-based LCPA but still achieves competing learning performance. [ABSTRACT FROM AUTHOR]

The implementation of artificial intelligence (AI) in wireless networks is becoming more and more popular because of the growing number of mobile devices and the availability of huge amount of data. However, directly transmitting data for centralized learning will cause long communication latency owing to the limited communication resource and may incur severe privacy issue as well. To address these issues, we consider the federated edge learning (FEEL) system in this paper and develop an importance-aware joint data selection and resource allocation algorithm to maximize the learning efficiency. Aiming at selecting important data for local training, we first analyze the relation between loss decay and gradient norm, which indicates that larger gradient norm generally leads to faster learning speed. Based on this, a learning efficiency maximization problem is formulated by jointly considering the communication resource allocation and data selection. The closed-form results for optimal communication resource allocation and data selection are both developed, where some insights are also highlighted. Furthermore, an optimal algorithm with low computational complexity is developed to obtain the optimal end-to-end latency in one training period. We show that the sample size should be set to its upper limit in order to maximize the learning performance. Finally, we conduct extensive experiments on three popular convolutional neural network (CNN) models. The results show that the proposed algorithm can effectively reduce the training latency and improve the learning accuracy as compared with some benchmark algorithms. [ABSTRACT FROM AUTHOR]

This paper proposes a novel approach for crowd counting in low to high density scenarios in static images. Current approaches cannot handle huge crowd diversity well and thus perform poorly in extreme cases, where the crowd density in different regions of an image is either too low or too high, leading to crowd underestimation or overestimation. The proposed solution is based on the observation that detecting and handling such extreme cases in a specialized way leads to better crowd estimation. Additionally, existing methods find it hard to differentiate between the actual crowd and the cluttered background regions, resulting in further count overestimation. To address these issues, we propose a simple yet effective modular approach, where an input image is first subdivided into fixed-size patches and then fed to a four-way classification module labeling each image patch as low, medium, high-dense or no-crowd. This module also provides a count for each label, which is then analyzed via a specifically devised novel decision module to decide whether the image belongs to any of the two extreme cases (very low or very high density) or a normal case. Images, specified as high- or low-density extreme or a normal case, pass through dedicated zooming or normal patch-making blocks respectively before routing to the regressor in the form of fixed-size patches for crowd estimate. Extensive experimental evaluations demonstrate that the proposed approach outperforms the state-of-the-art methods on four benchmarks under most of the evaluation criteria. [ABSTRACT FROM AUTHOR]

Histopathological image analysis is a challenging task due to a diverse histology feature set as well as due to the presence of large non-informative regions in whole slide images. In this paper, we propose a multiple-instance learning (MIL) method for image-level classification as well as for annotating relevant regions in the image. In MIL, a common assumption is that negative bags contain only negative instances while positive bags contain one or more positive instances. This asymmetric assumption may be inappropriate for some application scenarios where negative bags also contain representative negative instances. We introduce a novel symmetric MIL framework associating each instance in a bag with an attribute which can be either negative, positive, or irrelevant. We extend the notion of relevance by introducing control over the number of relevant instances. We develop a probabilistic graphical model that incorporates the aforementioned paradigm and a corresponding computationally efficient inference for learning the model parameters and obtaining an instance level attribute-learning classifier. The effectiveness of the proposed method is evaluated on available histopathology datasets with promising results. [ABSTRACT FROM AUTHOR]

Federated learning has emerged as an advanced privacy-preserving learning technique for mobile edge computing, where the model is trained in a decentralized manner by the clients, preventing the server from directly accessing those private data from the clients. This learning mechanism significantly challenges the attack from the server side. Although the state-of-the-art attacking techniques that incorporated the advance of Generative adversarial networks (GANs) could construct class representatives of the global data distribution among all clients, it is still challenging to distinguishably attack a specific client (i.e., user-level privacy leakage), which is a stronger privacy threat to precisely recover the private data from a specific client. To analyze the privacy leakage of federated learning, this paper gives the first attempt to explore user-level privacy leakage by the attack from a malicious server. We propose a framework incorporating GAN with a multi-task discriminator, called multi-task GAN – Auxiliary Identification (mGAN-AI), which simultaneously discriminates category, reality, and client identity of input samples. The novel discrimination on client identity enables the generator to recover user specified private data. Unlike existing works interfering the federated learning process, the proposed method works “invisibly” on the server side. Furthermore, considering the anonymization strategy for mitigating mGAN-AI, we propose a beforehand linkability attack which re-identifies the anonymized updates by associating the client representatives. A novel siamese network fusing the identification and verification models is developed for measuring the similarity of representatives. The experimental results demonstrate the effectiveness of the proposed approaches and the superior to the state-of-the-art. [ABSTRACT FROM AUTHOR]

This paper presents a novel model-driven deep learning (DL) architecture, called TurboNet, for turbo decoding that integrates DL into the traditional max-log-maximum a posteriori (MAP) algorithm. The TurboNet inherits the superiority of the max-log-MAP algorithm and DL tools and thus presents excellent error-correction capability with low training cost. To design the TurboNet, the original iterative structure is unfolded as deep neural network (DNN) decoding units, where trainable weights are introduced to the max-log-MAP algorithm and optimized through supervised learning. To efficiently train the TurboNet, a loss function is carefully designed to prevent tricky gradient vanishing issue. To further reduce the computational complexity and training cost of the TurboNet, we can prune it into TurboNet+. Compared with the existing black-box DL approaches, the TurboNet+ has considerable advantage in computational complexity and is conducive to significantly reducing the decoding overhead. Furthermore, we also present a simple training strategy to address the overfitting issue, which enable efficient training of the proposed TurboNet+. Simulation results demonstrate TurboNet+’s superiority in error-correction ability, signal-to-noise ratio generalization, and computational overhead. In addition, an experimental system is established for an over-the-air (OTA) test with the help of a 5G rapid prototyping system and demonstrates TurboNet’s strong learning ability and great robustness to various scenarios. [ABSTRACT FROM AUTHOR]

This paper introduces a framework for systematic complexity scaling of deep neural network (DNN) based MIMO detectors. The model uses a fraction of the DNN inputs by scaling their values through weights that follow monotonically non-increasing functions. This allows for weight scaling across and within the different DNN layers in order to achieve accuracy-vs.-complexity scalability during inference. In order to further improve the performance of our proposal, we introduce a sparsity-inducing regularization constraint in conjunction with trainable weight-scaling functions. In this way, the network learns to balance detection accuracy versus complexity while also increasing robustness to changes in the activation patterns, leading to further improvement in the detection accuracy and BER performance at the same inference complexity. Numerical results show that our approach is 10-fold and 100-fold less complex than classical approaches based on semi-definite relaxation and ML detection, respectively. [ABSTRACT FROM AUTHOR]

The thermal capacity of buildings enables heating, ventilating, and air-conditioning (HVAC) systems to be exploited as demand response (DR) resources. Optimal DR of HVAC units is challenging, particularly for multi-zone buildings, because this requires detailed physics-based models of zonal temperature variations for HVAC system operation and building thermal conditions. Using supervised learning (SL), this paper proposes a new strategy for optimal DR of an HVAC system in a multi-zone building. Artificial neural networks (ANNs) are trained with data obtained under normal building operating conditions. The ANNs are replicated using piecewise linear equations, which are explicitly integrated into an optimization problem for price-based DR. The problem is solved for various electricity prices and building thermal conditions. The solutions are used to train a deep neural network (DNN) to directly determine the optimal DR schedule, which is termed as SL-aided meta-prediction (SLAMP) here; the DNN can work as a price-and-optimal-demand curve. Case studies are performed using three different methods: explicit ANN replication, SLAMP, and physics-based modeling. The case study results verify that the proposed SL-based strategy is effective in terms of both practical applicability and computational time, while also ensuring the thermal comfort of occupants and cost-effective operation of the HVAC system. [ABSTRACT FROM AUTHOR]

With the development of high-resolution fingerprint scanners, high-resolution fingerprint-based biometric recognition has received increasing attention in recent years. This paper presents a pore feature-based approach for biometric recognition. Our approach employs a convolutional neural network (CNN) model, DeepResPore, to detect pores in the input fingerprint image. Thereafter, a CNN-based descriptor is computed for a patch around each detected pore. Specifically, we have designed a residual learning-based CNN, referred to as PoreNet that learns distinctive feature representation from pore patches. For verification, a matching score is generated by comparing the pore descriptors, obtained from a pair of fingerprint images, in a bi-directional manner using the Euclidean distance. The proposed approach for high-resolution fingerprint recognition achieves 2.27% and 0.24% equal error rates (EERs) on partial (DBI) and complete (DBII) fingerprints of the benchmark PolyU HRF dataset. Most importantly, it achieves lower FMR1000 and FMR10000 values than the current state-of-the-art approach on both the datasets. Further, this is the first study to report the performance of a learning-based fingerprint recognition approach on cross-sensor fingerprint images. [ABSTRACT FROM AUTHOR]

Wafer maps contain information about various defect patterns on the wafer surface and automatic classification of these defects plays a vital role to find their root causes. Semiconductor engineers apply various methods for wafer defect classification such as manual visual inspection or machine learning-based algorithms by manually extracting useful features. However, these methods are unreliable, and their classification performance is also poor. Therefore, this paper proposes a deep learning-based convolutional neural network for automatic wafer defect identification (CNN-WDI). We applied a data augmentation technique to overcome the class-imbalance issue. The proposed model uses convolution layers to extract valuable features instead of manual feature extraction. Moreover, state-of-the-art regularization methods such as batch normalization and spatial dropout are used to improve the classification performance of the CNN-WDI model. The experimental results comparison using a real wafer dataset shows that our model outperformed all previously proposed machine learning-based wafer defect classification models. The average classification accuracy of the CNN-WDI model with nine different wafer map defects is 96.2%, which is an increment of 6.4% from the last highest average accuracy using the same dataset. [ABSTRACT FROM AUTHOR]

Design and analysis of high-speed SerDes channels primarily deal with ensuring signal integrity (SI) for desired electrical performance. SI is predominantly judged by time domain (TD) metrics: bit error rate (BER), eye-height (EH), and eye-width (EW). With increasing bit rates and stringent BER criteria, TD simulations are becoming compute-time-intensive. A full-factorial, cost-effective design space exploration for SI is made possible by learning-based mapping of frequency-domain S-parameter data to EH/EW. A major challenge in this mapping procedure is the identification of relevant S-parameter data, such as return loss, insertion loss, crosstalks, and the frequency points at which they are sampled. This paper outlines a methodology to identify the critical S-parameters at specific frequency points using information theory-based definition of data relevance using a fast correlation-based filter solution for feature selection. This technique is applied for identifying relevant features for generating artificial neural network-based prediction models of EH/EW within 2.5% accuracy for channels with high data rates and complex topologies. [ABSTRACT FROM PUBLISHER]

As an interesting and emerging topic, co-saliency detection aims at simultaneously extracting common salient objects from a group of images. On one hand, traditional co-saliency detection approaches rely heavily on human knowledge for designing hand-crafted metrics to possibly reflect the faithful properties of the co-salient regions. Such strategies, however, always suffer from poor generalization capability to flexibly adapt various scenarios in real applications. On the other hand, most current methods pursue co-saliency detection in unsupervised fashions. This, however, tends to weaken their performance in real complex scenarios because they are lack of robust learning mechanism to make full use of the weak labels of each image. To alleviate these two problems, this paper proposes a new SP-MIL framework for co-saliency detection, which integrates both multiple instance learning (MIL) and self-paced learning (SPL) into a unified learning framework. Specifically, for the first problem, we formulate the co-saliency detection problem as a MIL paradigm to learn the discriminative classifiers to detect the co-saliency object in the “instance-level”. The formulated MIL component facilitates our method capable of automatically producing the proper metrics to measure the intra-image contrast and the inter-image consistency for detecting co-saliency in a purely self-learning way. For the second problem, the embedded SPL paradigm is able to alleviate the data ambiguity under the weak supervision of co-saliency detection and guide a robust learning manner in complex scenarios. Experiments on benchmark datasets together with multiple extended computer vision applications demonstrate the superiority of the proposed framework beyond the state-of-the-arts. [ABSTRACT FROM AUTHOR]

Probabilistic principal component regression (PPCR) has been introduced for soft sensor modeling as a probabilistic projection regression method, which is effective in handling data collinearity and random noises. However, the linear limitation of data relationships may cause its performance deterioration when applied to nonlinear processes. Therefore, a novel weighted PPCR (WPPCR) algorithm is proposed in this paper for soft sensing of nonlinear processes. In WPPCR, by including the most relevant samples for local modeling, different weights will be assigned to these samples according to their similarities with the testing sample. Then, a weighted log-likelihood function is constructed, and expectation-maximization algorithm can be carried out iteratively to obtain the optimal model parameters. In this way, the nonlinear data relationship can be locally approximated by WPPCR. The effectiveness and flexibility of the proposed method are validated on a numerical example and an industrial process. [ABSTRACT FROM AUTHOR]