July 28, 2019

3021 words 15 mins read

Paper Group ANR 386

Deep Belief Networks used on High Resolution Multichannel Electroencephalography Data for Seizure Detection. Measurement of amplitude of the moiré patterns in digital autostereoscopic 3D display. Who is Smarter? Intelligence Measure of Learning-based Cognitive Radios. Dynamic Deep Neural Networks: Optimizing Accuracy-Efficiency Trade-offs by Select …

Deep Belief Networks used on High Resolution Multichannel Electroencephalography Data for Seizure Detection


Title	Deep Belief Networks used on High Resolution Multichannel Electroencephalography Data for Seizure Detection
Authors	JT Turner, Adam Page, Tinoosh Mohsenin, Tim Oates
Abstract	Ubiquitous bio-sensing for personalized health monitoring is slowly becoming a reality with the increasing availability of small, diverse, robust, high fidelity sensors. This oncoming flood of data begs the question of how we will extract useful information from it. In this paper we explore the use of a variety of representations and machine learning algorithms applied to the task of seizure detection in high resolution, multichannel EEG data. We explore classification accuracy, computational complexity and memory requirements with a view toward understanding which approaches are most suitable for such tasks as the number of people involved and the amount of data they produce grows to be quite large. In particular, we show that layered learning approaches such as Deep Belief Networks excel along these dimensions.
Tasks	EEG, Seizure Detection
Published	2017-08-28
URL	http://arxiv.org/abs/1708.08430v1
PDF	http://arxiv.org/pdf/1708.08430v1.pdf
PWC	https://paperswithcode.com/paper/deep-belief-networks-used-on-high-resolution
Repo
Framework

Measurement of amplitude of the moiré patterns in digital autostereoscopic 3D display


Title	Measurement of amplitude of the moiré patterns in digital autostereoscopic 3D display
Authors	Vladimir Saveljev, Sung-Kyu Kim
Abstract	The article presents the experimental measurements of the amplitude of the moir'e patterns in a digital autostereoscopic barrier-type 3D display across a wide angular range with a small increment. The period and orientation of the moir'e patterns were also measured as functions of the angle. Simultaneous branches are observed and analyzed. The theoretical interpretation is also given. The results can help preventing or minimizing the moir'e effect in displays.
Tasks
Published	2017-09-18
URL	http://arxiv.org/abs/1709.07745v1
PDF	http://arxiv.org/pdf/1709.07745v1.pdf
PWC	https://paperswithcode.com/paper/measurement-of-amplitude-of-the-moire
Repo
Framework

Who is Smarter? Intelligence Measure of Learning-based Cognitive Radios


Title	Who is Smarter? Intelligence Measure of Learning-based Cognitive Radios
Authors	Monireh Dabaghchian, Amir Alipour-Fanid, Songsong Liu, Kai Zeng, Xiaohua Li, Yu Chen
Abstract	Cognitive radio (CR) is considered as a key enabling technology for dynamic spectrum access to improve spectrum efficiency. Although the CR concept was invented with the core idea of realizing cognition, the research on measuring CR cognitive capabilities and intelligence is largely open. Deriving the intelligence measure of CR not only can lead to the development of new CR technologies, but also makes it possible to better configure the networks by integrating CRs with different cognitive capabilities. In this paper, for the first time, we propose a data-driven methodology to quantitatively measure the intelligence factors of the CR with learning capabilities. The basic idea of our methodology is to run various tests on the CR in different spectrum environments under different settings and obtain various performance data on different metrics. Then we apply factor analysis on the performance data to identify and quantize the intelligence factors and cognitive capabilities of the CR. More specifically, we present a case study consisting of 144 different types of CRs. The CRs are different in terms of learning-based dynamic spectrum access strategies, number of sensors, sensing accuracy, processing speed, and algorithmic complexity. Five intelligence factors are identified for the CRs through our data analysis.We show that these factors comply well with the nature of the tested CRs, which validates the proposed intelligence measure methodology.
Tasks
Published	2017-12-26
URL	http://arxiv.org/abs/1712.09315v2
PDF	http://arxiv.org/pdf/1712.09315v2.pdf
PWC	https://paperswithcode.com/paper/who-is-smarter-intelligence-measure-of
Repo
Framework

Dynamic Deep Neural Networks: Optimizing Accuracy-Efficiency Trade-offs by Selective Execution


Title	Dynamic Deep Neural Networks: Optimizing Accuracy-Efficiency Trade-offs by Selective Execution
Authors	Lanlan Liu, Jia Deng
Abstract	We introduce Dynamic Deep Neural Networks (D2NN), a new type of feed-forward deep neural network that allows selective execution. Given an input, only a subset of D2NN neurons are executed, and the particular subset is determined by the D2NN itself. By pruning unnecessary computation depending on input, D2NNs provide a way to improve computational efficiency. To achieve dynamic selective execution, a D2NN augments a feed-forward deep neural network (directed acyclic graph of differentiable modules) with controller modules. Each controller module is a sub-network whose output is a decision that controls whether other modules can execute. A D2NN is trained end to end. Both regular and controller modules in a D2NN are learnable and are jointly trained to optimize both accuracy and efficiency. Such training is achieved by integrating backpropagation with reinforcement learning. With extensive experiments of various D2NN architectures on image classification tasks, we demonstrate that D2NNs are general and flexible, and can effectively optimize accuracy-efficiency trade-offs.
Tasks	Image Classification
Published	2017-01-02
URL	http://arxiv.org/abs/1701.00299v3
PDF	http://arxiv.org/pdf/1701.00299v3.pdf
PWC	https://paperswithcode.com/paper/dynamic-deep-neural-networks-optimizing
Repo
Framework

Learning Robot Activities from First-Person Human Videos Using Convolutional Future Regression


Title	Learning Robot Activities from First-Person Human Videos Using Convolutional Future Regression
Authors	Jangwon Lee, Michael S. Ryoo
Abstract	We design a new approach that allows robot learning of new activities from unlabeled human example videos. Given videos of humans executing the same activity from a human’s viewpoint (i.e., first-person videos), our objective is to make the robot learn the temporal structure of the activity as its future regression network, and learn to transfer such model for its own motor execution. We present a new deep learning model: We extend the state-of-the-art convolutional object detection network for the representation/estimation of human hands in training videos, and newly introduce the concept of using a fully convolutional network to regress (i.e., predict) the intermediate scene representation corresponding to the future frame (e.g., 1-2 seconds later). Combining these allows direct prediction of future locations of human hands and objects, which enables the robot to infer the motor control plan using our manipulation network. We experimentally confirm that our approach makes learning of robot activities from unlabeled human interaction videos possible, and demonstrate that our robot is able to execute the learned collaborative activities in real-time directly based on its camera input.
Tasks	Object Detection
Published	2017-03-03
URL	http://arxiv.org/abs/1703.01040v2
PDF	http://arxiv.org/pdf/1703.01040v2.pdf
PWC	https://paperswithcode.com/paper/learning-robot-activities-from-first-person
Repo
Framework

On the ability of neural nets to express distributions


Title	On the ability of neural nets to express distributions
Authors	Holden Lee, Rong Ge, Tengyu Ma, Andrej Risteski, Sanjeev Arora
Abstract	Deep neural nets have caused a revolution in many classification tasks. A related ongoing revolution—also theoretically not understood—concerns their ability to serve as generative models for complicated types of data such as images and texts. These models are trained using ideas like variational autoencoders and Generative Adversarial Networks. We take a first cut at explaining the expressivity of multilayer nets by giving a sufficient criterion for a function to be approximable by a neural network with $n$ hidden layers. A key ingredient is Barron’s Theorem \cite{Barron1993}, which gives a Fourier criterion for approximability of a function by a neural network with 1 hidden layer. We show that a composition of $n$ functions which satisfy certain Fourier conditions (“Barron functions”) can be approximated by a $n+1$-layer neural network. For probability distributions, this translates into a criterion for a probability distribution to be approximable in Wasserstein distance—a natural metric on probability distributions—by a neural network applied to a fixed base distribution (e.g., multivariate gaussian). Building up recent lower bound work, we also give an example function that shows that composition of Barron functions is more expressive than Barron functions alone.
Tasks
Published	2017-02-22
URL	http://arxiv.org/abs/1702.07028v2
PDF	http://arxiv.org/pdf/1702.07028v2.pdf
PWC	https://paperswithcode.com/paper/on-the-ability-of-neural-nets-to-express
Repo
Framework

Signal Recovery from Unlabeled Samples


Title	Signal Recovery from Unlabeled Samples
Authors	Saeid Haghighatshoar, Giuseppe Caire
Abstract	In this paper, we study the recovery of a signal from a set of noisy linear projections (measurements), when such projections are unlabeled, that is, the correspondence between the measurements and the set of projection vectors (i.e., the rows of the measurement matrix) is not known a priori. We consider a special case of unlabeled sensing referred to as Unlabeled Ordered Sampling (UOS) where the ordering of the measurements is preserved. We identify a natural duality between this problem and classical Compressed Sensing (CS), where we show that the unknown support (location of nonzero elements) of a sparse signal in CS corresponds to the unknown indices of the measurements in UOS. While in CS it is possible to recover a sparse signal from an under-determined set of linear equations (less equations than the signal dimension), successful recovery in UOS requires taking more samples than the dimension of the signal. Motivated by this duality, we develop a Restricted Isometry Property (RIP) similar to that in CS. We also design a low-complexity Alternating Minimization algorithm that achieves a stable signal recovery under the established RIP. We analyze our proposed algorithm for different signal dimensions and number of measurements theoretically and investigate its performance empirically via simulations. The results are reminiscent of phase-transition similar to that occurring in CS.
Tasks
Published	2017-01-30
URL	http://arxiv.org/abs/1701.08701v4
PDF	http://arxiv.org/pdf/1701.08701v4.pdf
PWC	https://paperswithcode.com/paper/signal-recovery-from-unlabeled-samples
Repo
Framework

Pose Estimation using Local Structure-Specific Shape and Appearance Context


Title	Pose Estimation using Local Structure-Specific Shape and Appearance Context
Authors	Anders Glent Buch, Dirk Kraft, Joni-Kristian Kamarainen, Henrik Gordon Petersen, Norbert Krüger
Abstract	We address the problem of estimating the alignment pose between two models using structure-specific local descriptors. Our descriptors are generated using a combination of 2D image data and 3D contextual shape data, resulting in a set of semi-local descriptors containing rich appearance and shape information for both edge and texture structures. This is achieved by defining feature space relations which describe the neighborhood of a descriptor. By quantitative evaluations, we show that our descriptors provide high discriminative power compared to state of the art approaches. In addition, we show how to utilize this for the estimation of the alignment pose between two point sets. We present experiments both in controlled and real-life scenarios to validate our approach.
Tasks	Pose Estimation
Published	2017-08-23
URL	http://arxiv.org/abs/1708.06963v1
PDF	http://arxiv.org/pdf/1708.06963v1.pdf
PWC	https://paperswithcode.com/paper/pose-estimation-using-local-structure
Repo
Framework

A global feature extraction model for the effective computer aided diagnosis of mild cognitive impairment using structural MRI images


Title	A global feature extraction model for the effective computer aided diagnosis of mild cognitive impairment using structural MRI images
Authors	Chen Fang, Panuwat Janwattanapong, Chunfei Li, Malek Adjouadi
Abstract	Multiple modalities of biomarkers have been proved to be very sensitive in assessing the progression of Alzheimer’s disease (AD), and using these modalities and machine learning algorithms, several approaches have been proposed to assist in the early diagnosis of AD. Among the recent investigated state-of-the-art approaches, Gaussian discriminant analysis (GDA)-based approaches have been demonstrated to be more effective and accurate in the classification of AD, especially for delineating its prodromal stage of mild cognitive impairment (MCI). Moreover, among those binary classification investigations, the local feature extraction methods were mostly used, which made them hardly be applied to a practical computer aided diagnosis system. Therefore, this study presents a novel global feature extraction model taking advantage of the recent proposed GDA-based dual high-dimensional decision spaces, which can significantly improve the early diagnosis performance comparing to those local feature extraction methods. In the true test using 20% held-out data, for discriminating the most challenging MCI group from the cognitively normal control (CN) group, an F1 score of 91.06%, an accuracy of 88.78%, a sensitivity of 91.80%, and a specificity of 83.78% were achieved that can be considered as the best performance obtained so far.
Tasks
Published	2017-12-02
URL	http://arxiv.org/abs/1712.00556v1
PDF	http://arxiv.org/pdf/1712.00556v1.pdf
PWC	https://paperswithcode.com/paper/a-global-feature-extraction-model-for-the
Repo
Framework


Title	Incremental Learning in Deep Convolutional Neural Networks Using Partial Network Sharing
Authors	Syed Shakib Sarwar, Aayush Ankit, Kaushik Roy
Abstract	Deep convolutional neural network (DCNN) based supervised learning is a widely practiced approach for large-scale image classification. However, retraining these large networks to accommodate new, previously unseen data demands high computational time and energy requirements. Also, previously seen training samples may not be available at the time of retraining. We propose an efficient training methodology and incrementally growing DCNN to learn new tasks while sharing part of the base network. Our proposed methodology is inspired by transfer learning techniques, although it does not forget previously learned tasks. An updated network for learning new set of classes is formed using previously learned convolutional layers (shared from initial part of base network) with addition of few newly added convolutional kernels included in the later layers of the network. We employed a `clone-and-branch’ technique which allows the network to learn new tasks one after another without any performance loss in old tasks. We evaluated the proposed scheme on several recognition applications. The classification accuracy achieved by our approach is comparable to the regular incremental learning approach (where networks are updated with new training samples only, without any network sharing), while achieving energy efficiency, reduction in storage requirements, memory access and training time. \|
Tasks	Image Classification, Transfer Learning
Published	2017-12-07
URL	http://arxiv.org/abs/1712.02719v4
PDF	http://arxiv.org/pdf/1712.02719v4.pdf
PWC	https://paperswithcode.com/paper/incremental-learning-in-deep-convolutional
Repo
Framework

Estimating the risk associated with transportation technology using multifidelity simulation


Title	Estimating the risk associated with transportation technology using multifidelity simulation
Authors	Erik J. Schlicht, Nichole L. Morris
Abstract	This paper provides a quantitative method for estimating the risk associated with candidate transportation technology, before it is developed and deployed. The proposed solution extends previous methods that rely exclusively on low-fidelity human-in-the-loop experimental data, or high-fidelity traffic data, by adopting a multifidelity approach that leverages data from both low- and high-fidelity sources. The multifidelity method overcomes limitations inherent to existing approaches by allowing a model to be trained inexpensively, while still assuring that its predictions generalize to the real-world. This allows for candidate technologies to be evaluated at the stage of conception, and enables a mechanism for only the safest and most effective technology to be developed and released.
Tasks
Published	2017-01-30
URL	http://arxiv.org/abs/1701.08588v2
PDF	http://arxiv.org/pdf/1701.08588v2.pdf
PWC	https://paperswithcode.com/paper/estimating-the-risk-associated-with
Repo
Framework

A Thorough Formalization of Conceptual Spaces


Title	A Thorough Formalization of Conceptual Spaces
Authors	Lucas Bechberger, Kai-Uwe Kühnberger
Abstract	The highly influential framework of conceptual spaces provides a geometric way of representing knowledge. Instances are represented by points in a high-dimensional space and concepts are represented by convex regions in this space. After pointing out a problem with the convexity requirement, we propose a formalization of conceptual spaces based on fuzzy star-shaped sets. Our formalization uses a parametric definition of concepts and extends the original framework by adding means to represent correlations between different domains in a geometric way. Moreover, we define computationally efficient operations on concepts (intersection, union, and projection onto a subspace) and show that these operations can support both learning and reasoning processes.
Tasks
Published	2017-06-20
URL	http://arxiv.org/abs/1706.06366v2
PDF	http://arxiv.org/pdf/1706.06366v2.pdf
PWC	https://paperswithcode.com/paper/a-thorough-formalization-of-conceptual-spaces
Repo
Framework

Neural Network-Based Automatic Liver Tumor Segmentation With Random Forest-Based Candidate Filtering


Title	Neural Network-Based Automatic Liver Tumor Segmentation With Random Forest-Based Candidate Filtering
Authors	Grzegorz Chlebus, Hans Meine, Jan Hendrik Moltz, Andrea Schenk
Abstract	We present a fully automatic method employing convolutional neural networks based on the 2D U-net architecture and random forest classifier to solve the automatic liver lesion segmentation problem of the ISBI 2017 Liver Tumor Segmentation Challenge (LiTS). In order to constrain the ROI in which the tumors could be located, a liver segmentation is performed first. For the organ segmentation, an ensemble of convolutional networks is trained to segment a liver using a set of 179 liver CT datasets from liver surgery planning. Inside of the liver ROI a neural network, trained using 127 challenge training datasets, identifies tumor candidates, which are subsequently filtered with a random forest classifier yielding the final tumor segmentation. The evaluation on the 70 challenge test cases resulted in a mean Dice coefficient of 0.65, ranking our method in the second place.
Tasks	Lesion Segmentation, Liver Segmentation
Published	2017-06-02
URL	http://arxiv.org/abs/1706.00842v3
PDF	http://arxiv.org/pdf/1706.00842v3.pdf
PWC	https://paperswithcode.com/paper/neural-network-based-automatic-liver-tumor
Repo
Framework

Language Modeling by Clustering with Word Embeddings for Text Readability Assessment


Title	Language Modeling by Clustering with Word Embeddings for Text Readability Assessment
Authors	Miriam Cha, Youngjune Gwon, H. T. Kung
Abstract	We present a clustering-based language model using word embeddings for text readability prediction. Presumably, an Euclidean semantic space hypothesis holds true for word embeddings whose training is done by observing word co-occurrences. We argue that clustering with word embeddings in the metric space should yield feature representations in a higher semantic space appropriate for text regression. Also, by representing features in terms of histograms, our approach can naturally address documents of varying lengths. An empirical evaluation using the Common Core Standards corpus reveals that the features formed on our clustering-based language model significantly improve the previously known results for the same corpus in readability prediction. We also evaluate the task of sentence matching based on semantic relatedness using the Wiki-SimpleWiki corpus and find that our features lead to superior matching performance.
Tasks	Language Modelling, Word Embeddings
Published	2017-09-05
URL	http://arxiv.org/abs/1709.01888v1
PDF	http://arxiv.org/pdf/1709.01888v1.pdf
PWC	https://paperswithcode.com/paper/language-modeling-by-clustering-with-word
Repo
Framework

Modeling Source Syntax for Neural Machine Translation


Title	Modeling Source Syntax for Neural Machine Translation
Authors	Junhui Li, Deyi Xiong, Zhaopeng Tu, Muhua Zhu, Min Zhang, Guodong Zhou
Abstract	Even though a linguistics-free sequence to sequence model in neural machine translation (NMT) has certain capability of implicitly learning syntactic information of source sentences, this paper shows that source syntax can be explicitly incorporated into NMT effectively to provide further improvements. Specifically, we linearize parse trees of source sentences to obtain structural label sequences. On the basis, we propose three different sorts of encoders to incorporate source syntax into NMT: 1) Parallel RNN encoder that learns word and label annotation vectors parallelly; 2) Hierarchical RNN encoder that learns word and label annotation vectors in a two-level hierarchy; and 3) Mixed RNN encoder that stitchingly learns word and label annotation vectors over sequences where words and labels are mixed. Experimentation on Chinese-to-English translation demonstrates that all the three proposed syntactic encoders are able to improve translation accuracy. It is interesting to note that the simplest RNN encoder, i.e., Mixed RNN encoder yields the best performance with an significant improvement of 1.4 BLEU points. Moreover, an in-depth analysis from several perspectives is provided to reveal how source syntax benefits NMT.
Tasks	Machine Translation
Published	2017-05-02
URL	http://arxiv.org/abs/1705.01020v1
PDF	http://arxiv.org/pdf/1705.01020v1.pdf
PWC	https://paperswithcode.com/paper/modeling-source-syntax-for-neural-machine
Repo
Framework