IT博客汇
  • 首页
  • 精华
  • 技术
  • 设计
  • 资讯
  • 扯淡
  • 权利声明
  • 登录 注册

    arXiv Paper Daily: Mon, 17 Oct 2016

    我爱机器学习(52ml.net)发表于 2016-10-17 00:00:00
    love 0

    Neural and Evolutionary Computing

    Hadamard Product for Low-rank Bilinear Pooling

    Jin-Hwa Kim, Kyoung Woon On, Jeonghee Kim, Jung-Woo Ha, Byoung-Tak Zhang
    Comments: 13 pages, 1 figure, & appendix included
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)

    Bilinear models provide rich representations compared to linear models. They
    have been applied in various visual tasks, such as object recognition,
    segmentation, and visual question-answering, to get state-of-the-art
    performances taking advantage of the expanded representations. However,
    bilinear representations tend to be high-dimensional, limiting the
    applicability to computationally complex tasks. We propose low-rank bilinear
    neural networks using Hadamard product (element-wise multiplication), commonly
    implemented in many scientific computing frameworks. We show that our model
    outperforms compact bilinear pooling in visual question-answering tasks, having
    a better parsimonious property.


    Computer Vision and Pattern Recognition

    Kernel Alignment Inspired Linear Discriminant Analysis

    Shuai Zheng, Chris Ding
    Comments: Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD, 2014
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)

    Kernel alignment measures the degree of similarity between two kernels. In
    this paper, inspired from kernel alignment, we propose a new Linear
    Discriminant Analysis (LDA) formulation, kernel alignment LDA (kaLDA). We first
    define two kernels, data kernel and class indicator kernel. The problem is to
    find a subspace to maximize the alignment between subspace-transformed data
    kernel and class indicator kernel. Surprisingly, the kernel alignment induced
    kaLDA objective function is very similar to classical LDA and can be expressed
    using between-class and total scatter matrices. This can be extended to
    multi-label data. We use a Stiefel-manifold gradient descent algorithm to solve
    this problem. We perform experiments on 8 single-label and 6 multi-label data
    sets. Results show that kaLDA has very good performance on many single-label
    and multi-label problems.

    Comparing Face Detection and Recognition Techniques

    Jyothi Korra
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    This paper implements and compares different techniques for face detection
    and recognition. One is find where the face is located in the images that is
    face detection and second is face recognition that is identifying the person.
    We study three techniques in this paper: Face detection using self organizing
    map (SOM), Face recognition by projection and nearest neighbor and Face
    recognition using SVM.

    Are Accuracy and Robustness Correlated?

    Andras Rozsa, Manuel Günther, Terrance E. Boult
    Comments: Submitted for review at ICMLA 2016
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Machine learning models are vulnerable to adversarial examples formed by
    applying small carefully chosen perturbations to inputs that cause unexpected
    classification errors. In this paper, we perform experiments on various
    adversarial example generation approaches with multiple deep convolutional
    neural networks including Residual Networks, the best performing models on
    ImageNet Large-Scale Visual Recognition Challenge 2015. We compare the
    adversarial example generation techniques with respect to the quality of the
    produced images, and measure the robustness of the tested machine learning
    models to adversarial examples. Finally, we conduct large-scale experiments on
    cross-model adversarial portability. We find that adversarial examples are
    mostly transferable across similar network topologies, and we demonstrate that
    better machine learning models are less vulnerable to adversarial examples.

    On Duality Of Multiple Target Tracking and Segmentation

    Yicong Tian, Mubarak Shah
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Traditionally, object tracking and segmentation are treated as two separate
    problems and solved independently. However, in this paper, we argue that
    tracking and segmentation are actually closely related and solving one should
    help the other. On one hand, the object track, which is a set of bounding boxes
    with one bounding box in every frame, would provide strong high-level guidance
    for the target/background segmentation task. On the other hand, the object
    segmentation would separate object from other objects and background, which
    will be useful for determining track locations in every frame. We propose a
    novel framework which combines online multiple target tracking and segmentation
    in a video. In our approach, the tracking and segmentation problems are coupled
    by Lagrange dual decomposition, which leads to more accurate segmentation
    results and also emph{helps resolve typical difficulties in multiple target
    tracking, such as occlusion handling, ID-switch and track drifting}. To track
    targets, an individual appearance model is learned for each target via
    structured learning and network flow is employed to generate tracks from
    densely sampled candidates. For segmentation, multi-label Conditional Random
    Field (CRF) is applied to a superpixel based spatio-temporal graph in a segment
    of video to assign background or target labels to every superpixel. The
    experiments on diverse sequences show that our method outperforms
    state-of-the-art approaches for multiple target tracking as well as
    segmentation.

    Amortised MAP Inference for Image Super-resolution

    Casper Kaae Sønderby, Jose Caballero, Lucas Theis, Wenzhe Shi, Ferenc Huszár
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Learning (cs.LG); Machine Learning (stat.ML)

    Image Super-resolution (SR) is an underdetermined inverse problem, where a
    large number of plausible high-resolution images can explain the same
    downsampled image. Most current single image SR methods use empirical risk
    minimisation, often with a pixel-wise mean squared error (MSE) loss. However,
    the outputs from such methods tend to be blurry, over-smoothed and generally
    appear implausible. A more desirable approach would employ Maximum a Posteriori
    (MAP) inference, preferring solutions that always have a high probability under
    the image prior, and thus appear more plausible. Direct MAP estimation for SR
    is non-trivial, as it requires us to build a model for the image prior from
    samples. Furthermore, MAP inference is often performed via optimisation-based
    iterative algorithms which don’t compare well with the efficiency of
    neural-network-based alternatives. Here we introduce new methods for amortised
    MAP inference whereby we calculate the MAP estimate directly using a
    convolutional neural network. We first introduce a novel neural network
    architecture that performs a projection to the affine subspace of valid SR
    solutions ensuring that the high resolution output of the network is always
    consistent with the low resolution input. We show that, using this
    architecture, the amortised MAP inference problem reduces to minimising the
    cross-entropy between two distributions, similar to training generative models.
    We propose three methods to solve this optimisation problem: (1) Generative
    Adversarial Networks (GAN) (2) denoiser-guided SR which backpropagates
    gradient-estimates from denoising to train the network, and (3) a baseline
    method using a maximum-likelihood-trained image prior. Our experiments show
    that the GAN based approach performs best on real image data, achieving
    particularly good results in photo-realistic texture SR.

    A Reduction Theorem for the Sample Mean in Dynamic Time Warping Spaces

    Brijnesh J. Jain, David Schultz
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Optimization and Control (math.OC); Machine Learning (stat.ML)

    Though the concept of sample mean in dynamic time warping (DTW) spaces is
    used in pattern recognition applications, its existence has neither been proved
    nor called into question. This article shows that a sample mean exists under
    general conditions that cover common variations of different DTW-spaces
    mentioned in the literature. The existence proofs are based on a Reduction
    Theorem that bounds the length of the candidate solutions we need to consider.
    The proposed results place the concept of sample mean in DTW-spaces on a sound
    mathematical foundation and serves as a first step towards a statistical theory
    of DTW-spaces.

    Hadamard Product for Low-rank Bilinear Pooling

    Jin-Hwa Kim, Kyoung Woon On, Jeonghee Kim, Jung-Woo Ha, Byoung-Tak Zhang
    Comments: 13 pages, 1 figure, & appendix included
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)

    Bilinear models provide rich representations compared to linear models. They
    have been applied in various visual tasks, such as object recognition,
    segmentation, and visual question-answering, to get state-of-the-art
    performances taking advantage of the expanded representations. However,
    bilinear representations tend to be high-dimensional, limiting the
    applicability to computationally complex tasks. We propose low-rank bilinear
    neural networks using Hadamard product (element-wise multiplication), commonly
    implemented in many scientific computing frameworks. We show that our model
    outperforms compact bilinear pooling in visual question-answering tasks, having
    a better parsimonious property.

    Learning and Fusing Multimodal Features from and for Multi-task Facial Computing

    Wei Li, Zhigang Zhu
    Comments: An experiment to feature fusion in deep learning
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    We propose a deep learning-based feature fusion approach for facial computing
    including face recognition as well as gender, race and age detection. Instead
    of training a single classifier on face images to classify them based on the
    features of the person whose face appears in the image, we first train four
    different classifiers for classifying face images based on race, age, gender
    and identification (ID). Multi-task features are then extracted from the
    trained models and cross-task-feature training is conducted which shows the
    value of fusing multimodal features extracted from multi-tasks. We have found
    that features trained for one task can be used for other related tasks. More
    interestingly, the features trained for a task with more classes (e.g. ID) and
    then used in another task with fewer classes (e.g. race) outperforms the
    features trained for the other task itself. The final feature fusion is
    performed by combining the four types of features extracted from the images by
    the four classifiers. The feature fusion approach improves the classifications
    accuracy by a 7.2%, 20.1%, 22.2%, 21.8% margin, respectively, for ID, age, race
    and gender recognition, over the results of single classifiers trained only on
    their individual features. The proposed method can be applied to applications
    in which different types of data or features can be extracted.

    Recurrent 3D Attentional Networks for End-to-End Active Object Recognition in Cluttered Scenes

    Min Liu, Yifei Shi, Lintao Zheng, Kai Xu
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Active vision is inherently attention-driven: The agent selects views of
    observation to best approach the vision task while improving its internal
    representation of the scene being observed. Inspired by the recent success of
    attention-based models in 2D vision tasks based on single RGB images, we
    propose to address the multi-view depth-based active object recognition using
    attention mechanism, through developing an end-to-end recurrent 3D attentional
    network. The architecture comprises of a recurrent neural network (RNN),
    storing and updating an internal representation, and two levels of spatial
    transformer units, guiding two-level attentions. Our model, trained with a 3D
    shape database, is able to iteratively attend to the best views targeting an
    object of interest for recognizing it, and focus on the object in each view for
    removing the background clutter. To realize 3D view selection, we derive a 3D
    spatial transformer network which is differentiable for training with
    back-propagation, achieving must faster convergence than the reinforcement
    learning employed by most existing attention-based models. Experiments show
    that our method outperforms state-of-the-art methods in cluttered scenes.

    Improved phase-unwrapping method using geometric constraints

    Guangliang Du, Minmin Wang, Canlin Zhou, Shuchun Si, Hui Li, Zhenkun Lei, Yanjie Li
    Comments: 15 pages, 11 figures
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Conventional dual-frequency fringe projection algorithm often suffers from
    phase unwrapping failure when the frequency ratio between the high frequency
    and the low one is too large. Zhang et.al. proposed an enhanced two-frequency
    phase-shifting method to use geometric constraints of digital fringe
    projection(DFP) to reduce the noise impact due to the large frequency ratio.
    However, this method needs to calibrate the DFP system and calculate the
    minimum phase map at the nearest position from the camera perspective, these
    procedures are are relatively complex and more time-cosuming. In this paper, we
    proposed an improved method, which eliminates the system calibration and
    determination in Zhang’s method,meanwhile does not need to use the low
    frequency fringe pattern. In the proposed method,we only need a set of high
    frequency fringe patterns to measure the object after the high frequency is
    directly estimated by the experiment. Thus the proposed method can simplify the
    procedure and improve the speed. Finally, the experimental evaluation is
    conducted to prove the validity of the proposed method.The results demonstrate
    that the proposed method can overcome the main disadvantages encountered by
    Zhang’s method.

    Assessing Threat of Adversarial Examples on Deep Neural Networks

    Abigail Graese, Andras Rozsa, Terrance E. Boult
    Comments: This is a pre-print version to appear in IEEE ICMLA 2016
    Subjects: Computer Vision and Pattern Recognition (cs.CV)

    Deep neural networks are facing a potential security threat from adversarial
    examples, inputs that look normal but cause an incorrect classification by the
    deep neural network. For example, the proposed threat could result in
    hand-written digits on a scanned check being incorrectly classified but looking
    normal when humans see them. This research assesses the extent to which
    adversarial examples pose a security threat, when one considers the normal
    image acquisition process. This process is mimicked by simulating the
    transformations that normally occur in acquiring the image in a real world
    application, such as using a scanner to acquire digits for a check amount or
    using a camera in an autonomous car. These small transformations negate the
    effect of the carefully crafted perturbations of adversarial examples,
    resulting in a correct classification by the deep neural network. Thus just
    acquiring the image decreases the potential impact of the proposed security
    threat. We also show that the already widely used process of averaging over
    multiple crops neutralizes most adversarial examples. Normal preprocessing,
    such as text binarization, almost completely neutralizes adversarial examples.
    This is the first paper to show that for text driven classification,
    adversarial examples are an academic curiosity, not a security threat.

    Message-passing algorithms for synchronization problems over compact groups

    Amelia Perry, Alexander S. Wein, Afonso S. Bandeira, Ankur Moitra
    Comments: 35 pages, 11 figures
    Subjects: Information Theory (cs.IT); Computer Vision and Pattern Recognition (cs.CV); Data Structures and Algorithms (cs.DS); Optimization and Control (math.OC); Machine Learning (stat.ML)

    Various alignment problems arising in cryo-electron microscopy, community
    detection, time synchronization, computer vision, and other fields fall into a
    common framework of synchronization problems over compact groups such as Z/L,
    U(1), or SO(3). The goal of such problems is to estimate an unknown vector of
    group elements given noisy relative observations. We present an efficient
    iterative algorithm to solve a large class of these problems, allowing for any
    compact group, with measurements on multiple ‘frequency channels’ (Fourier
    modes, or more generally, irreducible representations of the group). Our
    algorithm is a highly efficient iterative method following the blueprint of
    approximate message passing (AMP), which has recently arisen as a central
    technique for inference problems such as structured low-rank estimation and
    compressed sensing. We augment the standard ideas of AMP with ideas from
    representation theory so that the algorithm can work with distributions over
    compact groups. Using standard but non-rigorous methods from statistical
    physics we analyze the behavior of our algorithm on a Gaussian noise model,
    identifying phases where the problem is easy, (computationally) hard, and
    (statistically) impossible. In particular, such evidence predicts that our
    algorithm is information-theoretically optimal in many cases, and that the
    remaining cases show evidence of statistical-to-computational gaps.

    Automatically tracking neurons in a moving and deforming brain

    Jeffrey P. Nguyen, Ashley N. Linder, George S. Plummer, Joshua W. Shaevitz, Andrew M. Leifer
    Comments: 33 pages, 7 figures, code available
    Subjects: Neurons and Cognition (q-bio.NC); Computer Vision and Pattern Recognition (cs.CV); Biological Physics (physics.bio-ph)

    Advances in optical neuroimaging techniques now allow neural activity to be
    recorded with cellular resolution in awake and behaving animals. Brain motion
    in these recordings pose a unique challenge. The location of individual neurons
    must be tracked in 3D over time to accurately extract single neuron activity
    traces. Recordings from small invertebrates like C. elegans are especially
    challenging because they undergo very large brain motion and deformation during
    animal movement. Here we present an automated computer vision pipeline to
    reliably track populations of neurons with single neuron resolution in the
    brain of a freely moving C. elegans undergoing large motion and deformation. 3D
    volumetric fluorescent images of the animal’s brain are straightened, aligned
    and registered, and the locations of neurons in the images are found via
    segmentation. Each neuron is then assigned an identity using a new
    time-independent machine-learning approach we call Neuron Registration Vector
    Encoding. In this approach, non-rigid point-set registration is used to match
    each segmented neuron in each volume with a set of reference volumes taken from
    throughout the recording. The way each neuron matches with the references
    defines a feature vector which is clustered to assign an identity to each
    neuron in each volume. Finally, thin-plate spline interpolation is used to
    correct errors in segmentation and check consistency of assigned identities.
    The Neuron Registration Vector Encoding approach proposed here is uniquely well
    suited for tracking neurons in brains undergoing large deformations. When
    applied to whole-brain calcium imaging recordings in freely moving C. elegans,
    this analysis pipeline located 150 neurons for the duration of an 8 minute
    recording and consistently found more neurons more quickly than manual or
    semi-automated approaches.

    Generalization Error of Invariant Classifiers

    Jure Sokolic, Raja Giryes, Guillermo Sapiro, Miguel R. D. Rodrigues
    Comments: 9 pages, 3 figures
    Subjects: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Learning (cs.LG)

    This paper studies the generalization error of invariant classifiers. In
    particular, we consider the common scenario where the classification task is
    invariant to certain transformations of the input, and that the classifier is
    constructed (or learned) to be invariant to these transformations. Our approach
    relies on factoring the input space into a product of a base space and a set of
    transformations. We show that whereas the generalization error of a
    non-invariant classifier is proportional to the complexity of the input space,
    the generalization error of an invariant classifier is proportional to the
    complexity of the base space. We also derive a set of sufficient conditions on
    the geometry of the base space and the set of transformations that ensure that
    the complexity of the base space is much smaller than the complexity of the
    input space. Our analysis applies to general classifiers such as convolutional
    neural networks. We demonstrate the implications of the developed theory for
    such classifiers with experiments on the MNIST and CIFAR-10 datasets.

    Numerical Inversion of SRNF Maps for Elastic Shape Analysis of Genus-Zero Surfaces

    Hamid Laga, Qian Xie, Ian H. Jermyn, Anuj Srivastava
    Subjects: Graphics (cs.GR); Computational Geometry (cs.CG); Computer Vision and Pattern Recognition (cs.CV)

    Recent developments in elastic shape analysis (ESA) are motivated by the fact
    that it provides comprehensive frameworks for simultaneous registration,
    deformation, and comparison of shapes. These methods achieve computational
    efficiency using certain square-root representations that transform invariant
    elastic metrics into Euclidean metrics, allowing for applications of standard
    algorithms and statistical tools. For analyzing shapes of embeddings of
    $mathbb{S}^2$ in $mathbb{R}^3$, Jermyn et al. introduced square-root normal
    fields (SRNFs) that transformed an elastic metric, with desirable invariant
    properties, into the $mathbb{L}^2$ metric. These SRNFs are essentially surface
    normals scaled by square-roots of infinitesimal area elements. A critical need
    in shape analysis is to invert solutions (deformations, averages, modes of
    variations, etc) computed in the SRNF space, back to the original surface space
    for visualizations and inferences. Due to the lack of theory for understanding
    SRNFs maps and their inverses, we take a numerical approach and derive an
    efficient multiresolution algorithm, based on solving an optimization problem
    in the surface space, that estimates surfaces corresponding to given SRNFs.
    This solution is found effective, even for complex shapes, e.g. human bodies
    and animals, that undergo significant deformations including bending and
    stretching. Specifically, we use this inversion for computing elastic shape
    deformations, transferring deformations, summarizing shapes, and for finding
    modes of variability in a given collection, while simultaneously registering
    the surfaces. We demonstrate the proposed algorithms using a statistical
    analysis of human body shapes, classification of generic surfaces and analysis
    of brain structures.


    Artificial Intelligence

    Kernel Alignment Inspired Linear Discriminant Analysis

    Shuai Zheng, Chris Ding
    Comments: Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD, 2014
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)

    Kernel alignment measures the degree of similarity between two kernels. In
    this paper, inspired from kernel alignment, we propose a new Linear
    Discriminant Analysis (LDA) formulation, kernel alignment LDA (kaLDA). We first
    define two kernels, data kernel and class indicator kernel. The problem is to
    find a subspace to maximize the alignment between subspace-transformed data
    kernel and class indicator kernel. Surprisingly, the kernel alignment induced
    kaLDA objective function is very similar to classical LDA and can be expressed
    using between-class and total scatter matrices. This can be extended to
    multi-label data. We use a Stiefel-manifold gradient descent algorithm to solve
    this problem. We perform experiments on 8 single-label and 6 multi-label data
    sets. Results show that kaLDA has very good performance on many single-label
    and multi-label problems.

    Generalization Error of Invariant Classifiers

    Jure Sokolic, Raja Giryes, Guillermo Sapiro, Miguel R. D. Rodrigues
    Comments: 9 pages, 3 figures
    Subjects: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Learning (cs.LG)

    This paper studies the generalization error of invariant classifiers. In
    particular, we consider the common scenario where the classification task is
    invariant to certain transformations of the input, and that the classifier is
    constructed (or learned) to be invariant to these transformations. Our approach
    relies on factoring the input space into a product of a base space and a set of
    transformations. We show that whereas the generalization error of a
    non-invariant classifier is proportional to the complexity of the input space,
    the generalization error of an invariant classifier is proportional to the
    complexity of the base space. We also derive a set of sufficient conditions on
    the geometry of the base space and the set of transformations that ensure that
    the complexity of the base space is much smaller than the complexity of the
    input space. Our analysis applies to general classifiers such as convolutional
    neural networks. We demonstrate the implications of the developed theory for
    such classifiers with experiments on the MNIST and CIFAR-10 datasets.

    Localization for Wireless Sensor Networks: A Neural Network Approach

    Shiu Kumar, Ronesh Sharma, Edwin Vans
    Comments: 11 pages, 7 figures, 1 table, IJCNC
    Journal-ref: International Journal of Computer Networks & Communications
    (IJCNC), vol. 8, pp. 61-71, January 2016
    Subjects: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI)

    As Wireless Sensor Networks are penetrating into the industrial domain, many
    research opportunities are emerging. One such essential and challenging
    application is that of node localization. A feed-forward neural network based
    methodology is adopted in this paper. The Received Signal Strength Indicator
    (RSSI) values of the anchor node beacons are used. The number of anchor nodes
    and their configurations has an impact on the accuracy of the localization
    system, which is also addressed in this paper. Five different training
    algorithms are evaluated to find the training algorithm that gives the best
    result. The multi-layer Perceptron (MLP) neural network model was trained using
    Matlab. In order to evaluate the performance of the proposed method in real
    time, the model obtained was then implemented on the Arduino microcontroller.
    With four anchor nodes, an average 2D localization error of 0.2953 m has been
    achieved with a 12-12-2 neural network structure. The proposed method can also
    be implemented on any other embedded microcontroller system.

    Amortised MAP Inference for Image Super-resolution

    Casper Kaae Sønderby, Jose Caballero, Lucas Theis, Wenzhe Shi, Ferenc Huszár
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Learning (cs.LG); Machine Learning (stat.ML)

    Image Super-resolution (SR) is an underdetermined inverse problem, where a
    large number of plausible high-resolution images can explain the same
    downsampled image. Most current single image SR methods use empirical risk
    minimisation, often with a pixel-wise mean squared error (MSE) loss. However,
    the outputs from such methods tend to be blurry, over-smoothed and generally
    appear implausible. A more desirable approach would employ Maximum a Posteriori
    (MAP) inference, preferring solutions that always have a high probability under
    the image prior, and thus appear more plausible. Direct MAP estimation for SR
    is non-trivial, as it requires us to build a model for the image prior from
    samples. Furthermore, MAP inference is often performed via optimisation-based
    iterative algorithms which don’t compare well with the efficiency of
    neural-network-based alternatives. Here we introduce new methods for amortised
    MAP inference whereby we calculate the MAP estimate directly using a
    convolutional neural network. We first introduce a novel neural network
    architecture that performs a projection to the affine subspace of valid SR
    solutions ensuring that the high resolution output of the network is always
    consistent with the low resolution input. We show that, using this
    architecture, the amortised MAP inference problem reduces to minimising the
    cross-entropy between two distributions, similar to training generative models.
    We propose three methods to solve this optimisation problem: (1) Generative
    Adversarial Networks (GAN) (2) denoiser-guided SR which backpropagates
    gradient-estimates from denoising to train the network, and (3) a baseline
    method using a maximum-likelihood-trained image prior. Our experiments show
    that the GAN based approach performs best on real image data, achieving
    particularly good results in photo-realistic texture SR.

    Distributional Inclusion Hypothesis for Tensor-based Composition

    Dimitri Kartsaklis, Mehrnoosh Sadrzadeh
    Comments: To appear in COLING 2016
    Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)

    According to the distributional inclusion hypothesis, entailment between
    words can be measured via the feature inclusions of their distributional
    vectors. In recent work, we showed how this hypothesis can be extended from
    words to phrases and sentences in the setting of compositional distributional
    semantics. This paper focuses on inclusion properties of tensors; its main
    contribution is a theoretical and experimental analysis of how feature
    inclusion works in different concrete models of verb tensors. We present
    results for relational, Frobenius, projective, and holistic methods and compare
    them to the simple vector addition, multiplication, min, and max models. The
    degrees of entailment thus obtained are evaluated via a variety of existing
    word-based measures, such as Weed’s and Clarke’s, KL-divergence, APinc,
    balAPinc, and two of our previously proposed metrics at the phrase/sentence
    level. We perform experiments on three entailment datasets, investigating which
    version of tensor-based composition achieves the highest performance when
    combined with the sentence-level measures.

    Hadamard Product for Low-rank Bilinear Pooling

    Jin-Hwa Kim, Kyoung Woon On, Jeonghee Kim, Jung-Woo Ha, Byoung-Tak Zhang
    Comments: 13 pages, 1 figure, & appendix included
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)

    Bilinear models provide rich representations compared to linear models. They
    have been applied in various visual tasks, such as object recognition,
    segmentation, and visual question-answering, to get state-of-the-art
    performances taking advantage of the expanded representations. However,
    bilinear representations tend to be high-dimensional, limiting the
    applicability to computationally complex tasks. We propose low-rank bilinear
    neural networks using Hadamard product (element-wise multiplication), commonly
    implemented in many scientific computing frameworks. We show that our model
    outperforms compact bilinear pooling in visual question-answering tasks, having
    a better parsimonious property.


    Information Retrieval

    A Comprehensive Comparative Study of Word and Sentence Similarity Measures

    Issa Atoum, Ahmed Otoom, Narayanan Kulathuramaiyer
    Comments: 7 pages,4 figures
    Journal-ref: International Journal of Computer Applications,2016,135(1),
    Foundation of Computer Science (FCS), NY, USA
    Subjects: Information Retrieval (cs.IR); Computation and Language (cs.CL)

    Sentence similarity is considered the basis of many natural language tasks
    such as information retrieval, question answering and text summarization. The
    semantic meaning between compared text fragments is based on the words semantic
    features and their relationships. This article reviews a set of word and
    sentence similarity measures and compares them on benchmark datasets. On the
    studied datasets, results showed that hybrid semantic measures perform better
    than both knowledge and corpus based measures.

    Symmetric Private Information Retrieval For MDS Coded Distributed Storage

    Qiwen Wang, Mikael Skoglund
    Subjects: Information Theory (cs.IT); Cryptography and Security (cs.CR); Information Retrieval (cs.IR)

    A user wants to retrieve a file from a database without revealing the
    identity of the file retrieved at the database, which is known as the problem
    of private information retrieval (PIR). If it is further required that the user
    obtains no information about the database other than the desired file, the
    concept of symmetric private information retrieval (SPIR) is introduced to
    guarantee privacy for both parties. In this paper, the problem of SPIR is
    studied for a database stored among $N$ nodes in a distributed way, by using an
    $(N,M)$-MDS storage code. The information-theoretic capacity of SPIR, defined
    as the maximum number of symbols of the desired file retrieved per downloaded
    symbol, for the coded database is derived. It is shown that the SPIR capacity
    for coded database is $1-frac{M}{N}$, when the amount of the shared common
    randomness of distributed nodes (unavailable at the user) is at least
    $frac{M}{N-M}$ times the file size. Otherwise, the SPIR capacity for the coded
    database equals zero.


    Computation and Language

    Distributional Inclusion Hypothesis for Tensor-based Composition

    Dimitri Kartsaklis, Mehrnoosh Sadrzadeh
    Comments: To appear in COLING 2016
    Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)

    According to the distributional inclusion hypothesis, entailment between
    words can be measured via the feature inclusions of their distributional
    vectors. In recent work, we showed how this hypothesis can be extended from
    words to phrases and sentences in the setting of compositional distributional
    semantics. This paper focuses on inclusion properties of tensors; its main
    contribution is a theoretical and experimental analysis of how feature
    inclusion works in different concrete models of verb tensors. We present
    results for relational, Frobenius, projective, and holistic methods and compare
    them to the simple vector addition, multiplication, min, and max models. The
    degrees of entailment thus obtained are evaluated via a variety of existing
    word-based measures, such as Weed’s and Clarke’s, KL-divergence, APinc,
    balAPinc, and two of our previously proposed metrics at the phrase/sentence
    level. We perform experiments on three entailment datasets, investigating which
    version of tensor-based composition achieves the highest performance when
    combined with the sentence-level measures.

    Civique: Using Social Media to Detect Urban Emergencies

    Diptesh Kanojia, Vishwajeet Kumar, Krithi Ramamritham
    Comments: This paper was presented at Workshop on Social Data Analytics and Management (SoDAM 2016), at Very Large Databases (VLDB 2016), in September 2016
    Subjects: Computation and Language (cs.CL); Computers and Society (cs.CY); Social and Information Networks (cs.SI)

    We present the Civique system for emergency detection in urban areas by
    monitoring micro blogs like Tweets. The system detects emergency related
    events, and classifies them into appropriate categories like “fire”,
    “accident”, “earthquake”, etc. We demonstrate our ideas by classifying Twitter
    posts in real time, visualizing the ongoing event on a map interface and
    alerting users with options to contact relevant authorities, both online and
    offline. We evaluate our classifiers for both the steps, i.e., emergency
    detection and categorization, and obtain F-scores exceeding 70% and 90%,
    respectively. We demonstrate Civique using a web interface and on an Android
    application, in realtime, and show its use for both tweet detection and
    visualization.

    A Language-independent and Compositional Model for Personality Trait Recognition from Short Texts

    Fei Liu, Julien Perez, Scott Nowson
    Comments: 10 pages, 2 figures, 2 tables
    Subjects: Computation and Language (cs.CL); Machine Learning (stat.ML)

    Many methods have been used to recognize author personality traits from text,
    typically combining linguistic feature engineering with shallow learning
    models, e.g. linear regression or Support Vector Machines. This work uses
    deep-learning-based models and atomic features of text, the characters, to
    build hierarchical, vectorial word and sentence representations for trait
    inference. This method, applied to a corpus of tweets, shows state-of-the-art
    performance across five traits and three languages (English, Spanish and
    Italian) compared with prior work in author profiling. The results, supported
    by preliminary visualisation work, are encouraging for the ability to detect
    complex human traits.

    Fast, Scalable Phrase-Based SMT Decoding

    Hieu Hoang, Nicolay Bogoychev, Lane Schwartz, Marcin Junczys-Dowmunt
    Subjects: Computation and Language (cs.CL)

    The utilization of statistical machine translation (SMT) has grown enormously
    over the last decade, many using open-source software developed by the NLP
    community. As commercial use has increased, there is need for software that is
    optimized for commercial requirements, in particular, fast phrase-based
    decoding and more efficient utilization of modern multicore servers.

    In this paper we re-examine the major components of phrase-based decoding and
    decoder implementation with particular emphasis on speed and scalability on
    multicore machines. The result is a drop-in replacement for the Moses decoder
    which is up to fifteen times faster and scales monotonically with the number of
    cores.

    A Comprehensive Comparative Study of Word and Sentence Similarity Measures

    Issa Atoum, Ahmed Otoom, Narayanan Kulathuramaiyer
    Comments: 7 pages,4 figures
    Journal-ref: International Journal of Computer Applications,2016,135(1),
    Foundation of Computer Science (FCS), NY, USA
    Subjects: Information Retrieval (cs.IR); Computation and Language (cs.CL)

    Sentence similarity is considered the basis of many natural language tasks
    such as information retrieval, question answering and text summarization. The
    semantic meaning between compared text fragments is based on the words semantic
    features and their relationships. This article reviews a set of word and
    sentence similarity measures and compares them on benchmark datasets. On the
    studied datasets, results showed that hybrid semantic measures perform better
    than both knowledge and corpus based measures.


    Distributed, Parallel, and Cluster Computing

    Big Data Analytics in Cloud environment using Hadoop

    Mansaf Alam, Kashish Ara Shakil
    Comments: conference paper ICMPAS-2016
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    The Big Data management is a problem right now. The Big Data growth is very
    high. It is very difficult to manage due to various characteristics. This
    manuscript focuses on Big Data analytics in cloud environment using Hadoop. We
    have classified the Big Data according to its characteristics like Volume,
    Value, Variety and Velocity. We have made various nodes to process the data
    based on their volume, velocity, value and variety. In this work we have
    classify the input data and routed to various processing node. At the last
    after processing from each node, we can combine the output of all nodes to get
    the final result. We have used Hadoop to partition the data as well as process
    it.

    Reproducible Experiments for Comparing Apache Flink and Apache Spark on Public Clouds

    Shelan Perera, Ashansa Perera, Kamal Hakimzadeh
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    Big data processing is a hot topic in today’s computer science world. There
    is a significant demand for analysing big data to satisfy many requirements of
    many industries. Emergence of the Kappa architecture created a strong
    requirement for a highly capable and efficient data processing engine.
    Therefore data processing engines such as Apache Flink and Apache Spark emerged
    in open source world to fulfil that efficient and high performing data
    processing requirement. There are many available benchmarks to evaluate those
    two data processing engines. But complex deployment patterns and dependencies
    make those benchmarks very difficult to reproduce by our own. This project has
    two main goals. They are making few of community accepted benchmarks easily
    reproducible on cloud and validate the performance claimed by those studies.

    A Quantitative Model for Predicting Cross-application Interference in Virtual Environments

    Maicon Melo Alves, Lúcia Maria de Assumpção Drummond
    Comments: 14 pages, 6 figures, 6 tables
    Subjects: Distributed, Parallel, and Cluster Computing (cs.DC)

    Cross-application interference can affect drastically performance of HPC
    applications when running in clouds. This problem is caused by concurrent
    access performed by co-located applications to shared and non-sliceable
    resources such as cache and memory. In order to address this issue, some works
    adopted a qualitative approach that does not take into account the amount of
    access to shared resources. In addition, a few works, even considering the
    amount of access, evaluated just the SLLC access contention as the root of this
    problem. However, our experiments revealed that interference is intrinsically
    related to the amount of simultaneous access to shared resources, besides
    showing that another shared resources, apart from SLLC, can also influence the
    interference suffered by co-located applications. In this paper, we present a
    quantitative model for predicting cross-application interference in virtual
    environments. Our proposed model takes into account the amount of simultaneous
    access to SLLC, DRAM and virtual network, and the similarity of application’s
    access burden to predict the level of interference suffered by applications
    when co-located in a same physical machine. Experiments considering a real
    petroleum reservoir simulator and applications from HPCC benchmark showed that
    our model reached an average and maximum prediction errors around 4\% and 12\%,
    besides achieving an error less than 10\% in approximately 96\% of all tested
    cases.


    Learning

    Data-Driven Threshold Machine: Scan Statistics, Change-Point Detection, and Extreme Bandits

    Shuang Li, Yao Xie, Le Song
    Comments: Submitted to conference
    Subjects: Learning (cs.LG); Statistics Theory (math.ST); Machine Learning (stat.ML)

    We present a novel distribution-free approach, the data-driven threshold
    machine (DTM), for a fundamental problem at the core of many learning tasks:
    choose a threshold for a given pre-specified level that bounds the tail
    probability of the maximum of a (possibly dependent but stationary) random
    sequence. We do not assume data distribution, but rather relying on the
    asymptotic distribution of extremal values, and reduce the problem to estimate
    three parameters of the extreme value distributions and the extremal index. We
    specially take care of data dependence via estimating extremal index since in
    many settings, such as scan statistics, change-point detection, and extreme
    bandits, where dependence in the sequence of statistics can be significant. Key
    features of our DTM also include robustness and the computational efficiency,
    and it only requires one sample path to form a reliable estimate of the
    threshold, in contrast to the Monte Carlo sampling approach which requires
    drawing a large number of sample paths. We demonstrate the good performance of
    DTM via numerical examples in various dependent settings.

    Improved Strongly Adaptive Online Learning using Coin Betting

    Kwang-Sung Jun, Francesco Orabona, Rebecca Willett, Stephen Wright
    Subjects: Machine Learning (stat.ML); Learning (cs.LG)

    This paper describes a new parameter-free online learning algorithm for
    changing environments. In comparing against algorithms with the same time
    complexity as ours, we obtain a strongly adaptive regret bound that is a factor
    of at least $sqrt{log(T)}$ better, where $T$ is the time horizon. Empirical
    results show that our algorithm outperforms state-of-the-art methods in
    learning with expert advice and metric learning scenarios.

    Generalization Error of Invariant Classifiers

    Jure Sokolic, Raja Giryes, Guillermo Sapiro, Miguel R. D. Rodrigues
    Comments: 9 pages, 3 figures
    Subjects: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Learning (cs.LG)

    This paper studies the generalization error of invariant classifiers. In
    particular, we consider the common scenario where the classification task is
    invariant to certain transformations of the input, and that the classifier is
    constructed (or learned) to be invariant to these transformations. Our approach
    relies on factoring the input space into a product of a base space and a set of
    transformations. We show that whereas the generalization error of a
    non-invariant classifier is proportional to the complexity of the input space,
    the generalization error of an invariant classifier is proportional to the
    complexity of the base space. We also derive a set of sufficient conditions on
    the geometry of the base space and the set of transformations that ensure that
    the complexity of the base space is much smaller than the complexity of the
    input space. Our analysis applies to general classifiers such as convolutional
    neural networks. We demonstrate the implications of the developed theory for
    such classifiers with experiments on the MNIST and CIFAR-10 datasets.

    The End of Optimism? An Asymptotic Analysis of Finite-Armed Linear Bandits

    Tor Lattimore, Csaba Szepesvari
    Comments: 13 pages
    Subjects: Machine Learning (stat.ML); Learning (cs.LG)

    Stochastic linear bandits are a natural and simple generalisation of
    finite-armed bandits with numerous practical applications. Current approaches
    focus on generalising existing techniques for finite-armed bandits, notably the
    optimism principle and Thompson sampling. While prior work has mostly been in
    the worst-case setting, we analyse the asymptotic instance-dependent regret and
    show matching upper and lower bounds on what is achievable. Surprisingly, our
    results show that no algorithm based on optimism or Thompson sampling will ever
    achieve the optimal rate, and indeed, can be arbitrarily far from optimal, even
    in very simple cases. This is a disturbing result because these techniques are
    standard tools that are widely used for sequential optimisation. For example,
    for generalised linear bandits and reinforcement learning.

    Amortised MAP Inference for Image Super-resolution

    Casper Kaae Sønderby, Jose Caballero, Lucas Theis, Wenzhe Shi, Ferenc Huszár
    Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Learning (cs.LG); Machine Learning (stat.ML)

    Image Super-resolution (SR) is an underdetermined inverse problem, where a
    large number of plausible high-resolution images can explain the same
    downsampled image. Most current single image SR methods use empirical risk
    minimisation, often with a pixel-wise mean squared error (MSE) loss. However,
    the outputs from such methods tend to be blurry, over-smoothed and generally
    appear implausible. A more desirable approach would employ Maximum a Posteriori
    (MAP) inference, preferring solutions that always have a high probability under
    the image prior, and thus appear more plausible. Direct MAP estimation for SR
    is non-trivial, as it requires us to build a model for the image prior from
    samples. Furthermore, MAP inference is often performed via optimisation-based
    iterative algorithms which don’t compare well with the efficiency of
    neural-network-based alternatives. Here we introduce new methods for amortised
    MAP inference whereby we calculate the MAP estimate directly using a
    convolutional neural network. We first introduce a novel neural network
    architecture that performs a projection to the affine subspace of valid SR
    solutions ensuring that the high resolution output of the network is always
    consistent with the low resolution input. We show that, using this
    architecture, the amortised MAP inference problem reduces to minimising the
    cross-entropy between two distributions, similar to training generative models.
    We propose three methods to solve this optimisation problem: (1) Generative
    Adversarial Networks (GAN) (2) denoiser-guided SR which backpropagates
    gradient-estimates from denoising to train the network, and (3) a baseline
    method using a maximum-likelihood-trained image prior. Our experiments show
    that the GAN based approach performs best on real image data, achieving
    particularly good results in photo-realistic texture SR.

    Theoretical Analysis of Domain Adaptation with Optimal Transport

    Ievgen Redko, Amaury Habrard, Marc Sebban
    Subjects: Machine Learning (stat.ML); Learning (cs.LG)

    Domain adaptation (DA) is an important and emerging field of machine learning
    that tackles the problem occurring when the distributions of training (source
    domain) and test (target domain) data are similar but different. Current
    theoretical results show that the efficiency of DA algorithms depends on their
    capacity of minimizing the divergence between source and target probability
    distributions. In this paper, we provide a theoretical study on the advantages
    that concepts borrowed from optimal transportation theory can bring to DA. In
    particular, we show that the Wasserstein metric can be used as a divergence
    measure between distributions to obtain generalization guarantees for three
    different learning settings: (i) classic DA with unsupervised target data (ii)
    DA combining source and target labeled data, (iii) multiple source DA. Based on
    the obtained results, we provide some insights showing when this analysis can
    be tighter than other existing frameworks. We also show that in the context of
    multiple source DA, the problem of estimating of the best joint hypothesis
    between source and target labeling functions can be reformulated using a
    Wasserstein distance-based loss function. We think that these results open the
    door to novel ideas and directions for DA.

    Semi-supervised Graph Embedding Approach to Dynamic Link Prediction

    Ryohei Hisano
    Subjects: Machine Learning (stat.ML); Learning (cs.LG); Social and Information Networks (cs.SI); Physics and Society (physics.soc-ph)

    We propose a simple discrete time semi-supervised graph embedding approach to
    link prediction in dynamic networks. The learned embedding reflects information
    from both the temporal and cross-sectional network structures, which is
    performed by defining the loss function as a weighted sum of the supervised
    loss from past dynamics and the unsupervised loss of predicting the
    neighborhood context in the current network. Our model is also capable of
    learning different embeddings for both formation and dissolution dynamics.
    These key aspects contributes to the predictive performance of our model and we
    provide experiments with three real–world dynamic networks showing that our
    method is comparable to state of the art methods in link formation prediction
    and outperforms state of the art baseline methods in link dissolution
    prediction.

    Spectral Inference Methods on Sparse Graphs: Theory and Applications

    Alaa Saade
    Comments: PhD dissertation
    Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Information Theory (cs.IT); Learning (cs.LG)

    In an era of unprecedented deluge of (mostly unstructured) data, graphs are
    proving more and more useful, across the sciences, as a flexible abstraction to
    capture complex relationships between complex objects. One of the main
    challenges arising in the study of such networks is the inference of
    macroscopic, large-scale properties affecting a large number of objects, based
    solely on the microscopic interactions between their elementary constituents.
    Statistical physics, precisely created to recover the macroscopic laws of
    thermodynamics from an idealized model of interacting particles, provides
    significant insight to tackle such complex networks.

    In this dissertation, we use methods derived from the statistical physics of
    disordered systems to design and study new algorithms for inference on graphs.
    Our focus is on spectral methods, based on certain eigenvectors of carefully
    chosen matrices, and sparse graphs, containing only a small amount of
    information. We develop an original theory of spectral inference based on a
    relaxation of various mean-field free energy optimizations. Our approach is
    therefore fully probabilistic, and contrasts with more traditional motivations
    based on the optimization of a cost function. We illustrate the efficiency of
    our approach on various problems, including community detection, randomized
    similarity-based clustering, and matrix completion.

    MML is not consistent for Neyman-Scott

    Michael Brand
    Comments: 50 pages, 0 figures
    Subjects: Machine Learning (stat.ML); Learning (cs.LG); Statistics Theory (math.ST)

    Minimum Message Length (MML) is a popular method for statistical inference,
    belonging to the Minimum Description Length (MDL) family. It is a general name
    for any of several computationally-feasible approximations to the generally
    NP-Hard Strict Minimum Message Length (SMML) estimator. One often-cited
    showcase for the power of MML is the Neyman-Scott estimation problem, where
    most popular estimation algorithms fail to produce a consistent result. MML’s
    performance on Neyman-Scott was analysed by Dowe and Wallace (1997) and by
    Wallace (2005) and MML was shown to be consistent for the problem. However,
    this analysis was not performed on SMML, but rather on two SMML approximations:
    Wallace-Freeman and Ideal Group. As for most estimation problems, the exact
    SMML solution is not known for Neyman-Scott. We analyse the Dowe-Wallace
    solution, and show that it hinges critically on the use of an unnatural prior
    for the problem. We argue that the Jeffreys prior is a more natural prior to
    assume in this case. Re-analysing the problem over its Jeffreys prior, we show
    that both the Ideal Group and the Wallace-Freeman approximations converge to
    the (inconsistent) Maximum Likelihood (ML) solution. We develop novel
    techniques that enable determining properties of the SMML estimator for some
    general families of estimation problems without requiring a full construction
    of the estimator, and use these to show that for many problems, including
    Neyman-Scott, the SMML estimator is not a point-estimator at all. Rather, it
    maps each observation to an entire continuum of estimates. Furthermore, using
    the tools developed we show that for Neyman-Scott the SMML estimate is
    inconsistent for all parameter sets as well as asymptotically. We discuss
    methodological problems in the arguments put forward by previous authors, who
    argued that MML is consistent for Neyman-Scott and in general.

    Approximate Counting, the Lovasz Local Lemma and Inference in Graphical Models

    Ankur Moitra
    Comments: 27 pages, 2 figures
    Subjects: Data Structures and Algorithms (cs.DS); Computational Complexity (cs.CC); Learning (cs.LG)

    In this paper we introduce a new approach for approximately counting in
    bounded degree systems with higher-order constraints. Our main result is an
    algorithm to approximately count the number of solutions to a CNF formula
    $Phi$ with at least $k$ variables per clause and degree at most $d$ when $k$
    is logarithmic in $d$. This closes an exponential gap between the known upper
    and lower bounds.

    Moreover our algorithm extends straightforwardly to approximate sampling,
    which shows that under Lovasz Local Lemma-like conditions it is not only
    possible to find a satisfying assignment, it is also possible to generate one
    approximately uniformly at random from the set of all satisfying assignments.
    Our approach is a significant departure from earlier techniques in approximate
    counting, and is based on a framework to bootstrap an oracle for computing
    marginal probabilities on individual variables. Finally, we give an application
    of our results to show that it is algorithmically possible to sample from the
    posterior distribution in an interesting class of graphical models.

    Sim-to-Real Robot Learning from Pixels with Progressive Nets

    Andrei A. Rusu, Matej Vecerik, Thomas Rothörl, Nicolas Heess, Razvan Pascanu, Raia Hadsell
    Subjects: Robotics (cs.RO); Learning (cs.LG)

    Applying end-to-end learning to solve complex, interactive, pixel-driven
    control tasks on a robot is an unsolved problem. Deep Reinforcement Learning
    algorithms are too slow to achieve performance on a real robot, but their
    potential has been demonstrated in simulated environments. We propose using
    progressive networks to bridge the reality gap and transfer learned policies
    from simulation to the real world. The progressive net approach is a general
    framework that enables reuse of everything from low-level visual features to
    high-level policies for transfer to new tasks, enabling a compositional, yet
    simple, approach to building complex skills. We present an early demonstration
    of this approach with a number of experiments in the domain of robot
    manipulation that focus on bridging the reality gap. Unlike other proposed
    approaches, our real-world experiments demonstrate successful task learning
    from raw visual input on a fully actuated robot manipulator. Moreover, rather
    than relying on model-based trajectory optimisation, the task learning is
    accomplished using only deep reinforcement learning and sparse rewards.


    Information Theory

    Message-passing algorithms for synchronization problems over compact groups

    Amelia Perry, Alexander S. Wein, Afonso S. Bandeira, Ankur Moitra
    Comments: 35 pages, 11 figures
    Subjects: Information Theory (cs.IT); Computer Vision and Pattern Recognition (cs.CV); Data Structures and Algorithms (cs.DS); Optimization and Control (math.OC); Machine Learning (stat.ML)

    Various alignment problems arising in cryo-electron microscopy, community
    detection, time synchronization, computer vision, and other fields fall into a
    common framework of synchronization problems over compact groups such as Z/L,
    U(1), or SO(3). The goal of such problems is to estimate an unknown vector of
    group elements given noisy relative observations. We present an efficient
    iterative algorithm to solve a large class of these problems, allowing for any
    compact group, with measurements on multiple ‘frequency channels’ (Fourier
    modes, or more generally, irreducible representations of the group). Our
    algorithm is a highly efficient iterative method following the blueprint of
    approximate message passing (AMP), which has recently arisen as a central
    technique for inference problems such as structured low-rank estimation and
    compressed sensing. We augment the standard ideas of AMP with ideas from
    representation theory so that the algorithm can work with distributions over
    compact groups. Using standard but non-rigorous methods from statistical
    physics we analyze the behavior of our algorithm on a Gaussian noise model,
    identifying phases where the problem is easy, (computationally) hard, and
    (statistically) impossible. In particular, such evidence predicts that our
    algorithm is information-theoretically optimal in many cases, and that the
    remaining cases show evidence of statistical-to-computational gaps.

    Symmetric Private Information Retrieval For MDS Coded Distributed Storage

    Qiwen Wang, Mikael Skoglund
    Subjects: Information Theory (cs.IT); Cryptography and Security (cs.CR); Information Retrieval (cs.IR)

    A user wants to retrieve a file from a database without revealing the
    identity of the file retrieved at the database, which is known as the problem
    of private information retrieval (PIR). If it is further required that the user
    obtains no information about the database other than the desired file, the
    concept of symmetric private information retrieval (SPIR) is introduced to
    guarantee privacy for both parties. In this paper, the problem of SPIR is
    studied for a database stored among $N$ nodes in a distributed way, by using an
    $(N,M)$-MDS storage code. The information-theoretic capacity of SPIR, defined
    as the maximum number of symbols of the desired file retrieved per downloaded
    symbol, for the coded database is derived. It is shown that the SPIR capacity
    for coded database is $1-frac{M}{N}$, when the amount of the shared common
    randomness of distributed nodes (unavailable at the user) is at least
    $frac{M}{N-M}$ times the file size. Otherwise, the SPIR capacity for the coded
    database equals zero.

    Capacity of Clustered Distributed Storage

    Jy-yong Sohn, Beongjun Choi, Sung Whan Yoon, Jaekyun Moon
    Comments: 7 pages, submitted to IEEE ICC 2017
    Subjects: Information Theory (cs.IT)

    A new system model reflecting the clustered structure of distributed storage
    is suggested to investigate bandwidth requirements for repairing failed storage
    nodes. Large data centers with multiple racks/disks or local networks of
    storage devices (e.g. sensor network) are good applications of the suggested
    clustered model. In realistic scenarios involving clustered storage structures,
    repairing storage nodes using intact nodes residing in other clusters is more
    bandwidth-consuming than restoring nodes based on information from
    intra-cluster nodes. Therefore, it is important to differentiate between
    intra-cluster repair bandwidth and cross-cluster repair bandwidth in modeling
    distributed storage. Capacity of the suggested model is obtained as a function
    of fundamental resources of distributed storage systems, namely, storage
    capacity, intra-cluster repair bandwidth and cross-cluster repair bandwidth.
    Based on the capacity expression, feasible sets of required resources which
    enable reliable storage are analyzed. It is shown that the cross-cluster
    traffic can be minimized to zero (i.e., intra-cluster local repair becomes
    possible) by allowing extra resources on storage capacity and intra-cluster
    repair bandwidth, according to a law specified in a closed-form. Moreover,
    trade-off between cross-cluster traffic and intra-cluster traffic is observed
    for sufficiently large storage capacity.

    A Geometrical-Statistical approach to outlier removal for TDOA measuments

    Marco Compagnoni, Alessia Pini, Antonio Canclini, Paolo Bestagini, Fabio Antonacci, Stefano Tubaro, Augusto Sarti
    Comments: 26 pages, 10 figure, 2 tables
    Subjects: Information Theory (cs.IT); Sound (cs.SD); Methodology (stat.ME)

    The curse of outlier measurements in estimation problems is a well known
    issue in a variety of fields. Therefore, outlier removal procedures, which
    enables the identification of spurious measurements within a set, have been
    developed for many different scenarios and applications. In this paper, we
    propose a statistically motivated outlier removal algorithm for time
    differences of arrival (TDOAs), or equivalently range differences (RD),
    acquired at sensor arrays. The method exploits the TDOA-space formalism and
    works by only knowing the relative sensor positions. As the proposed method is
    completely independent from the application for which measurements are used, it
    can be reliably used to identify outliers within a set of TDOA/RD measurements
    in different fields (e.g. acoustic source localization, sensor synchronization,
    radar, remote sensing, etc.). The whole theoretical derivation is validated by
    means of synthetic simulations and real experiments.

    On Two Conjectures about Permutation Trinomials over $mathbb{F}_{3^{2k}}$

    Nian Li
    Subjects: Information Theory (cs.IT)

    Permutation polynomials with few terms attracts researchers’ interest in
    recent years due to their simple algebraic form and some additional
    extraordinary properties. In this paper, by analyzing the quadratic factors of
    a fifth-degree polynomial and a seventh-degree polynomial over the finite field
    $mathbb{F}_{3^{2k}}$, two conjectures on permutation trinomials over
    $mathbb{F}_{3^{2k}}$ proposed recently by Li, Qu, Li and Fu are settled, where
    $k$ is a positive integer.

    Joint Source, Channel and Space-time Coding of Progressive Sources in MIMO Systems

    Meesue Shin, Laura Toni, Sang-Hyo Kim, Seok-Ho Chang
    Comments: 31 pages, 6 figures (15 subfigures), Submitted to IEEE Transactions on Image Processing
    Subjects: Information Theory (cs.IT)

    The optimization of joint source and channel coding for a sequence of
    numerous progressive packets is a challenging problem. Further, the problem
    becomes more complicated if the space-time coding is also involved with the
    optimization in a multiple-input multiple-output (MIMO) system. This is because
    the number of ways of jointly assigning channels codes and space-time codes to
    progressive packets is much larger than that of solely assigning channel codes
    to the packets. We are unaware of any feasible and complete solution for such
    optimization of joint source, channel, and space-time coding of progressive
    packets. This paper applies a parametric approach to address that complex joint
    optimization problem in a MIMO system. We use the parametric methodology to
    derive some useful theoretical results, and then exploit those results to
    propose an optimization method where the joint assignment of channel codes and
    space-time codes to the packets can be optimized in a packet-by-packet manner.
    As a result, the computational complexity of the optimization is exponentially
    reduced, compared to the conventional exhaustive search. The numerical results
    show that the proposed method significantly improves the peak-signal-to-noise
    ratio performance of the rate-based optimal solution in a MIMO system.

    Delay Minimization in Real-time Communications with Joint Buffering and Coding

    Jesper H. Sørensen, Petar Popovski, Jan Østergaard
    Subjects: Information Theory (cs.IT)

    We present a closed-form expression for the minimal delay that is achievable
    in a setting that combines a buffer and an erasure code, used to mitigate the
    packet delay variance. The erasure code is modeled according to the recent
    information-theoretic results on finite block length codes. Evaluations reveal
    that accurate knowledge of the network parameters is essential for optimal
    operation. Moreover, it is shown that, when the network packet delay variance
    is large, the buffer delay becomes negligible. Therefore, in this case the
    delay budget should be spent mainly on the erasure code.

    SNR Maximization as a Non-Linear Programming -Towards Optimal Spreading Sequence-

    Hirofumi Tsuda, Ken Umeno
    Comments: 7 pages IEEE ICC
    Subjects: Information Theory (cs.IT)

    Signal to Noise Ratio (SNR) is an important index for wireless
    communications. There are many methods for increasing SNR. In CDMA systems,
    spreading sequences are used. We consider the frequency-selective
    wide-sense-stationary uncorrelated-scattering (WSSUS) channel and evaluate the
    worst case of SNR. We construct the non-linear programing for maximizing the
    lower bound of the average of SNR. This problem becomes the convex programming
    if we did not take into account the norm constraint. We derive necessary
    conditions for optimal spreading sequences for the problem.

    Generalized and Extended Product Codes

    Mario Blaum, Steve Hetzler
    Comments: 37 pages
    Subjects: Information Theory (cs.IT)

    Generalized Product (GPC) Codes, an unification of Product Codes and
    Integrated Interleaved (II) Codes, are presented. Applications for approaches
    requiring local and global parities are described. The more general problem of
    extending product codes by adding global parities is studied and an upper bound
    on the minimum distance of such codes is obtained. Codes with one, two and
    three global parities whose minimum distances meet the bound are presented.
    Tradeoffs between optimality and field size are discussed.

    Tonal consonance parameters expose a hidden order in music

    Jorge Useche, Rafael Hurtado
    Comments: 15 pages, 12 figures, 4 tables
    Subjects: Sound (cs.SD); Information Theory (cs.IT); Data Analysis, Statistics and Probability (physics.data-an); Physics and Society (physics.soc-ph)

    Consonance is related to the perception of pleasantness arising from the
    combination of sounds and has been approached quantitatively using mathematical
    relations, physics, information theory and psychoacoustics. Tonal consonance is
    present in timbre, musical tuning, harmony and melody, and it is used for
    conveying sensations and emotions in music. It involves the physical properties
    of sound waves and is used to study melody and harmony through musical
    intervals and chords. From the perspective of complexity, the macroscopic
    properties of a system with many parts frequently rely on statistical
    properties of its constituent elements. Here we show that melody in musical
    pieces can be described in terms of the physical properties of melodic
    intervals and the existence of an entropy extremalization principle in music
    with psychoacoustic macroscopic constraints given by conserved quantities with
    musical meaning. This result connects human perception with human creativity
    through the physical properties of the musical stimulus.

    On Communication Cost vs. Load Balancing in Content Delivery Networks

    Mahdi Jafari Siavoshani, Seyed Pooya Shariatpanahi, Hamid Ghasemi, Ali Pourmiri
    Comments: 6 pages, 7 figures, submitted to conference
    Subjects: Networking and Internet Architecture (cs.NI); Information Theory (cs.IT)

    It is well known that load balancing and low delivery communication cost are
    two critical issues in mapping requests to servers in Content Delivery Networks
    (CDNs). However, the trade-off between these two performance metrics has not
    been yet quantitatively investigated in designing efficient request mapping
    schemes. In this work, we formalize this trade-off through a stochastic
    optimization problem. While the solutions to the problem in the extreme cases
    of minimum communication cost and optimum load balancing can be derived in
    closed form, finding the general solution is hard to derive. Thus we propose
    three heuristic mapping schemes and compare the trade-off performance of them
    through extensive simulations.

    Our simulation results show that at the expense of high query cost, we can
    achieve a good trade-off curve. Moreover, by benefiting from the power of
    multiple choices phenomenon, we can achieve almost the same performance with
    much less query cost. Finally, we can handle requests with different delay
    requirements at the cost of degrading network performance.

    Spectral Inference Methods on Sparse Graphs: Theory and Applications

    Alaa Saade
    Comments: PhD dissertation
    Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Information Theory (cs.IT); Learning (cs.LG)

    In an era of unprecedented deluge of (mostly unstructured) data, graphs are
    proving more and more useful, across the sciences, as a flexible abstraction to
    capture complex relationships between complex objects. One of the main
    challenges arising in the study of such networks is the inference of
    macroscopic, large-scale properties affecting a large number of objects, based
    solely on the microscopic interactions between their elementary constituents.
    Statistical physics, precisely created to recover the macroscopic laws of
    thermodynamics from an idealized model of interacting particles, provides
    significant insight to tackle such complex networks.

    In this dissertation, we use methods derived from the statistical physics of
    disordered systems to design and study new algorithms for inference on graphs.
    Our focus is on spectral methods, based on certain eigenvectors of carefully
    chosen matrices, and sparse graphs, containing only a small amount of
    information. We develop an original theory of spectral inference based on a
    relaxation of various mean-field free energy optimizations. Our approach is
    therefore fully probabilistic, and contrasts with more traditional motivations
    based on the optimization of a cost function. We illustrate the efficiency of
    our approach on various problems, including community detection, randomized
    similarity-based clustering, and matrix completion.

    Source localization and denoising: a perspective from the TDOA space

    Marco Compagnoni, Antonio Canclini, Paolo Bestagini, Fabio Antonacci, Augusto Sarti, Stefano Tubaro
    Comments: 25 pages, 9 figures
    Journal-ref: Multidimensional Systems and Signal Processing 2016
    Subjects: Sound (cs.SD); Information Theory (cs.IT); Methodology (stat.ME)

    In this manuscript, we formulate the problem of denoising Time Differences of
    Arrival (TDOAs) in the TDOA space, i.e. the Euclidean space spanned by TDOA
    measurements. The method consists of pre-processing the TDOAs with the purpose
    of reducing the measurement noise. The complete set of TDOAs (i.e., TDOAs
    computed at all microphone pairs) is known to form a redundant set, which lies
    on a linear subspace in the TDOA space. Noise, however, prevents TDOAs from
    lying exactly on this subspace. We therefore show that TDOA denoising can be
    seen as a projection operation that suppresses the component of the noise that
    is orthogonal to that linear subspace. We then generalize the projection
    operator also to the cases where the set of TDOAs is incomplete. We
    analytically show that this operator improves the localization accuracy, and we
    further confirm that via simulation.




沪ICP备19023445号-2号
友情链接