January 31, 2020

2939 words 14 mins read

Paper Group ANR 123

Generating Minimal Adversarial Perturbations with Integrated Adaptive Gradients. Polynomial Matrix Completion for Missing Data Imputation and Transductive Learning. Video Motion Capture from the Part Confidence Maps of Multi-Camera Images by Spatiotemporal Filtering Using the Human Skeletal Model. A Knowledge Graph-based Approach for Exploring the …

Generating Minimal Adversarial Perturbations with Integrated Adaptive Gradients


Title	Generating Minimal Adversarial Perturbations with Integrated Adaptive Gradients
Authors	Yatie Xiao, Chi-Man Pun
Abstract	Deep neural networks are easily fooled high confidence predictions for adversarial samples
Tasks	Image Classification
Published	2019-04-12
URL	https://arxiv.org/abs/1904.06186v3
PDF	https://arxiv.org/pdf/1904.06186v3.pdf
PWC	https://paperswithcode.com/paper/generating-minimal-adversarial-perturbations
Repo
Framework

Polynomial Matrix Completion for Missing Data Imputation and Transductive Learning


Title	Polynomial Matrix Completion for Missing Data Imputation and Transductive Learning
Authors	Jicong Fan, Yuqian Zhang, Madeleine Udell
Abstract	This paper develops new methods to recover the missing entries of a high-rank or even full-rank matrix when the intrinsic dimension of the data is low compared to the ambient dimension. Specifically, we assume that the columns of a matrix are generated by polynomials acting on a low-dimensional intrinsic variable, and wish to recover the missing entries under this assumption. We show that we can identify the complete matrix of minimum intrinsic dimension by minimizing the rank of the matrix in a high dimensional feature space. We develop a new formulation of the resulting problem using the kernel trick together with a new relaxation of the rank objective, and propose an efficient optimization method. We also show how to use our methods to complete data drawn from multiple nonlinear manifolds. Comparative studies on synthetic data, subspace clustering with missing data, motion capture data recovery, and transductive learning verify the superiority of our methods over the state-of-the-art.
Tasks	Imputation, Matrix Completion, Motion Capture
Published	2019-12-15
URL	https://arxiv.org/abs/1912.06989v1
PDF	https://arxiv.org/pdf/1912.06989v1.pdf
PWC	https://paperswithcode.com/paper/polynomial-matrix-completion-for-missing-data
Repo
Framework

Video Motion Capture from the Part Confidence Maps of Multi-Camera Images by Spatiotemporal Filtering Using the Human Skeletal Model


Title	Video Motion Capture from the Part Confidence Maps of Multi-Camera Images by Spatiotemporal Filtering Using the Human Skeletal Model
Authors	Takuya Ohashi, Yosuke Ikegami, Kazuki Yamamoto, Wataru Takano, Yoshihiko Nakamura
Abstract	This paper discusses video motion capture, namely, 3D reconstruction of human motion from multi-camera images. After the Part Confidence Maps are computed from each camera image, the proposed spatiotemporal filter is applied to deliver the human motion data with accuracy and smoothness for human motion analysis. The spatiotemporal filter uses the human skeleton and mixes temporal smoothing in two-time inverse kinematics computations. The experimental results show that the mean per joint position error was 26.1mm for regular motions and 38.8mm for inverted motions.
Tasks	3D Reconstruction, Motion Capture
Published	2019-12-09
URL	https://arxiv.org/abs/1912.03880v2
PDF	https://arxiv.org/pdf/1912.03880v2.pdf
PWC	https://paperswithcode.com/paper/video-motion-capture-from-the-part-confidence
Repo
Framework

A Knowledge Graph-based Approach for Exploring the U.S. Opioid Epidemic


Title	A Knowledge Graph-based Approach for Exploring the U.S. Opioid Epidemic
Authors	Maulik R. Kamdar, Tymor Hamamsy, Shea Shelton, Ayin Vala, Tome Eftimov, James Zou, Suzanne Tamang
Abstract	The United States is in the midst of an opioid epidemic with recent estimates indicating that more than 130 people die every day due to drug overdose. The over-prescription and addiction to opioid painkillers, heroin, and synthetic opioids, has led to a public health crisis and created a huge social and economic burden. Statistical learning methods that use data from multiple clinical centers across the US to detect opioid over-prescribing trends and predict possible opioid misuse are required. However, the semantic heterogeneity in the representation of clinical data across different centers makes the development and evaluation of such methods difficult and non-trivial. We create the Opioid Drug Knowledge Graph (ODKG) – a network of opioid-related drugs, active ingredients, formulations, combinations, and brand names. We use the ODKG to normalize drug strings in a clinical data warehouse consisting of patient data from over 400 healthcare facilities in 42 different states. We showcase the use of ODKG to generate summary statistics of opioid prescription trends across US regions. These methods and resources can aid the development of advanced and scalable models to monitor the opioid epidemic and to detect illicit opioid misuse behavior. Our work is relevant to policymakers and pain researchers who wish to systematically assess factors that contribute to opioid over-prescribing and iatrogenic opioid addiction in the US.
Tasks
Published	2019-05-27
URL	https://arxiv.org/abs/1905.11513v1
PDF	https://arxiv.org/pdf/1905.11513v1.pdf
PWC	https://paperswithcode.com/paper/a-knowledge-graph-based-approach-for
Repo
Framework

Fault Detection and Identification using Bayesian Recurrent Neural Networks


Title	Fault Detection and Identification using Bayesian Recurrent Neural Networks
Authors	Weike Sun, Antonio R. C. Paiva, Peng Xu, Anantha Sundaram, Richard D. Braatz
Abstract	In processing and manufacturing industries, there has been a large push to produce higher quality products and ensure maximum efficiency of processes. This requires approaches to effectively detect and resolve disturbances to ensure optimal operations. While the control system can compensate for many types of disturbances, there are changes to the process which it still cannot handle adequately. It is therefore important to further develop monitoring systems to effectively detect and identify those faults such that they can be quickly resolved by operators. In this paper, a novel probabilistic fault detection and identification method is proposed which adopts a newly developed deep learning approach using Bayesian recurrent neural networks (BRNNs) with variational dropout. The BRNN model is general and can model complex nonlinear dynamics. Moreover, compared to traditional statistic-based data-driven fault detection and identification methods, the proposed BRNN-based method yields uncertainty estimates which allow for simultaneous fault detection of chemical processes, direct fault identification, and fault propagation analysis. The outstanding performance of this method is demonstrated and contrasted to (dynamic) principal component analysis, which are widely applied in the industry, in the benchmark Tennessee Eastman process (TEP) and a real chemical manufacturing dataset.
Tasks	Fault Detection
Published	2019-11-11
URL	https://arxiv.org/abs/1911.04386v1
PDF	https://arxiv.org/pdf/1911.04386v1.pdf
PWC	https://paperswithcode.com/paper/fault-detection-and-identification-using
Repo
Framework

Mixture-based Multiple Imputation Model for Clinical Data with a Temporal Dimension


Title	Mixture-based Multiple Imputation Model for Clinical Data with a Temporal Dimension
Authors	Ye Xue, Diego Klabjan, Yuan Luo
Abstract	The problem of missing values in multivariable time series is a key challenge in many applications such as clinical data mining. Although many imputation methods show their effectiveness in many applications, few of them are designed to accommodate clinical multivariable time series. In this work, we propose a multiple imputation model that capture both cross-sectional information and temporal correlations. We integrate Gaussian processes with mixture models and introduce individualized mixing weights to handle the variance of predictive confidence of Gaussian process models. The proposed model is compared with several state-of-the-art imputation algorithms on both real-world and synthetic datasets. Experiments show that our best model can provide more accurate imputation than the benchmarks on all of our datasets.
Tasks	Gaussian Processes, Imputation, Time Series
Published	2019-08-12
URL	https://arxiv.org/abs/1908.04209v3
PDF	https://arxiv.org/pdf/1908.04209v3.pdf
PWC	https://paperswithcode.com/paper/mixture-based-multiple-imputation-models-for
Repo
Framework

SLTR: Simultaneous Localization of Target and Reflector in NLOS Condition Using Beacons


Title	SLTR: Simultaneous Localization of Target and Reflector in NLOS Condition Using Beacons
Authors	Muhammad. H Fares, Hadi Moradi, Mahmoud Shahabadi
Abstract	When the direct view between the target and the observer is not available, due to obstacles with non-zero sizes, the observation is received after reflection from a reflector, this is the indirect view or Non-Line-Of Sight condition. Localization of a target in NLOS condition still one of the open problems yet. In this paper, we address this problem by localizing the reflector and the target simultaneously using a single stationary receiver, and a determined number of beacons, in which their placements are also analyzed in an unknown map. The work is done in mirror space, when the receiver is a camera, and the reflector is a planar mirror. Furthermore, the distance from the observer to the target is estimated by size constancy concept, and the angle of coming signal is the same as the orientation of the camera, with respect to a global frame. The results show the validation of the proposed work and the simulation results are matched with the theoretical results.
Tasks
Published	2019-11-10
URL	https://arxiv.org/abs/1911.03940v1
PDF	https://arxiv.org/pdf/1911.03940v1.pdf
PWC	https://paperswithcode.com/paper/sltr-simultaneous-localization-of-target-and
Repo
Framework

On the Vector Space in Photoplethysmography Imaging


Title	On the Vector Space in Photoplethysmography Imaging
Authors	Christian S. Pilz, Vladimir Blazek, Steffen Leonhardt
Abstract	We study the vector space of visible wavelength intensities from face videos widely used as input features in Photoplethysmography Imaging (PPGI). Based upon theoretical principles of Group invariance in the Euclidean space we derive a change of the topology where the corresponding distance between successive measurements is defined as geodesic on a Riemannian manifold. This lower dimensional embedding of the sensor signal unifies the invariance properties with respect to translation of the features as discussed by several former approaches. The resulting operator acts implicit on the feature space without requiring any kind of prior knowledge and does not need parameter tuning. The resulting feature’s time varying quasi-periodic shaping naturally occurs in form of the canonical state space representation according to the known Diffusion process of blood volume changes. The computational complexity is low and the implementation becomes fairly simple. During experiments the operator achieved robust and competitive estimation performance of heart rate from face videos on two public databases.
Tasks
Published	2019-06-11
URL	https://arxiv.org/abs/1906.04431v1
PDF	https://arxiv.org/pdf/1906.04431v1.pdf
PWC	https://paperswithcode.com/paper/on-the-vector-space-in-photoplethysmography
Repo
Framework

Convolution, attention and structure embedding


Title	Convolution, attention and structure embedding
Authors	Jean-Marc Andreoli
Abstract	Deep neural networks are composed of layers of parametrised linear operations intertwined with non linear activations. In basic models, such as the multi-layer perceptron, a linear layer operates on a simple input vector embedding of the instance being processed, and produces an output vector embedding by straight multiplication by a matrix parameter. In more complex models, the input and output are structured and their embeddings are higher order tensors. The parameter of each linear operation must then be controlled so as not to explode with the complexity of the structures involved. This is essentially the role of convolution models, which exist in many flavours dependent on the type of structure they deal with (grids, networks, time series etc.). We present here a unified framework which aims at capturing the essence of these diverse models, allowing a systematic analysis of their properties and their mutual enrichment. We also show that attention models naturally fit in the same framework: attention is convolution in which the structure itself is adaptive, and learnt, instead of being given a priori.
Tasks	Time Series
Published	2019-05-03
URL	https://arxiv.org/abs/1905.01289v5
PDF	https://arxiv.org/pdf/1905.01289v5.pdf
PWC	https://paperswithcode.com/paper/convolution-is-outer-product
Repo
Framework

End-to-End Learning-Based Ultrasound Reconstruction


Title	End-to-End Learning-Based Ultrasound Reconstruction
Authors	Walter Simson, Rüdiger Göbl, Magdalini Paschali, Markus Krönke, Klemens Scheidhauer, Wolfgang Weber, Nassir Navab
Abstract	Ultrasound imaging is caught between the quest for the highest image quality, and the necessity for clinical usability. Our contribution is two-fold: First, we propose a novel fully convolutional neural network for ultrasound reconstruction. Second, a custom loss function tailored to the modality is employed for end-to-end training of the network. We demonstrate that training a network to map time-delayed raw data to a minimum variance ground truth offers performance increases in a clinical environment. In doing so, a path is explored towards improved clinically viable ultrasound reconstruction. The proposed method displays both promising image reconstruction quality and acquisition frequency when integrated for live ultrasound scanning. A clinical evaluation is conducted to verify the diagnostic usefulness of the proposed method in a clinical setting.
Tasks	Image Reconstruction
Published	2019-04-09
URL	http://arxiv.org/abs/1904.04696v1
PDF	http://arxiv.org/pdf/1904.04696v1.pdf
PWC	https://paperswithcode.com/paper/end-to-end-learning-based-ultrasound
Repo
Framework

Reactive Multi-Stage Feature Fusion for Multimodal Dialogue Modeling


Title	Reactive Multi-Stage Feature Fusion for Multimodal Dialogue Modeling
Authors	Yi-Ting Yeh, Tzu-Chuan Lin, Hsiao-Hua Cheng, Yu-Hsuan Deng, Shang-Yu Su, Yun-Nung Chen
Abstract	Visual question answering and visual dialogue tasks have been increasingly studied in the multimodal field towards more practical real-world scenarios. A more challenging task, audio visual scene-aware dialogue (AVSD), is proposed to further advance the technologies that connect audio, vision, and language, which introduces temporal video information and dialogue interactions between a questioner and an answerer. This paper proposes an intuitive mechanism that fuses features and attention in multiple stages in order to well integrate multimodal features, and the results demonstrate its capability in the experiments. Also, we apply several state-of-the-art models in other tasks to the AVSD task, and further analyze their generalization across different tasks.
Tasks	Question Answering, Scene-Aware Dialogue, Visual Dialog, Visual Question Answering
Published	2019-08-14
URL	https://arxiv.org/abs/1908.05067v1
PDF	https://arxiv.org/pdf/1908.05067v1.pdf
PWC	https://paperswithcode.com/paper/reactive-multi-stage-feature-fusion-for
Repo
Framework

Probabilistically-autoencoded horseshoe-disentangled multidomain item-response theory models


Title	Probabilistically-autoencoded horseshoe-disentangled multidomain item-response theory models
Authors	Joshua C. Chang, Shashaank Vattikuti, Carson C. Chow
Abstract	Item response theory (IRT) is a non-linear generative probabilistic paradigm for using exams to identify, quantify, and compare latent traits of individuals, relative to their peers, within a population of interest. In pre-existing multidimensional IRT methods, one requires a factorization of the test items. For this task, linear exploratory factor analysis is used, making IRT a posthoc model. We propose skipping the initial factor analysis by using a sparsity-promoting horseshoe prior to perform factorization directly within the IRT model so that all training occurs in a single self-consistent step. Being a hierarchical Bayesian model, we adapt the WAIC to the problem of dimensionality selection. IRT models are analogous to probabilistic autoencoders. By binding the generative IRT model to a Bayesian neural network (forming a probabilistic autoencoder), one obtains a scoring algorithm consistent with the interpretable Bayesian model. In some IRT applications the black-box nature of a neural network scoring machine is desirable. In this manuscript, we demonstrate within-IRT factorization and comment on scoring approaches.
Tasks
Published	2019-12-05
URL	https://arxiv.org/abs/1912.02351v1
PDF	https://arxiv.org/pdf/1912.02351v1.pdf
PWC	https://paperswithcode.com/paper/probabilistically-autoencoded-horseshoe
Repo
Framework

The Windfall Clause: Distributing the Benefits of AI for the Common Good


Title	The Windfall Clause: Distributing the Benefits of AI for the Common Good
Authors	Cullen O’Keefe, Peter Cihon, Ben Garfinkel, Carrick Flynn, Jade Leung, Allan Dafoe
Abstract	As the transformative potential of AI has become increasingly salient as a matter of public and political interest, there has been growing discussion about the need to ensure that AI broadly benefits humanity. This in turn has spurred debate on the social responsibilities of large technology companies to serve the interests of society at large. In response, ethical principles and codes of conduct have been proposed to meet the escalating demand for this responsibility to be taken seriously. As yet, however, few institutional innovations have been suggested to translate this responsibility into legal commitments which apply to companies positioned to reap large financial gains from the development and use of AI. This paper offers one potentially attractive tool for addressing such issues: the Windfall Clause, which is an ex ante commitment by AI firms to donate a significant amount of any eventual extremely large profits. By this we mean an early commitment that profits that a firm could not earn without achieving fundamental, economically transformative breakthroughs in AI capabilities will be donated to benefit humanity broadly, with particular attention towards mitigating any downsides from deployment of windfall-generating AI.
Tasks
Published	2019-12-25
URL	https://arxiv.org/abs/1912.11595v2
PDF	https://arxiv.org/pdf/1912.11595v2.pdf
PWC	https://paperswithcode.com/paper/the-windfall-clause-distributing-the-benefits
Repo
Framework

Cascaded Algorithm-Selection and Hyper-Parameter Optimization with Extreme-Region Upper Confidence Bound Bandit


Title	Cascaded Algorithm-Selection and Hyper-Parameter Optimization with Extreme-Region Upper Confidence Bound Bandit
Authors	Yi-Qi Hu, Yang Yu, Jun-Da Liao
Abstract	An automatic machine learning (AutoML) task is to select the best algorithm and its hyper-parameters simultaneously. Previously, the hyper-parameters of all algorithms are joint as a single search space, which is not only huge but also redundant, because many dimensions of hyper-parameters are irrelevant with the selected algorithms. In this paper, we propose a cascaded approach for algorithm selection and hyper-parameter optimization. While a search procedure is employed at the level of hyper-parameter optimization, a bandit strategy runs at the level of algorithm selection to allocate the budget based on the search feedbacks. Since the bandit is required to select the algorithm with the maximum performance, instead of the average performance, we thus propose the extreme-region upper confidence bound (ER-UCB) strategy, which focuses on the extreme region of the underlying feedback distribution. We show theoretically that the ER-UCB has a regret upper bound $O\left(K \ln n\right)$ with independent feedbacks, which is as efficient as the classical UCB bandit. We also conduct experiments on a synthetic problem as well as a set of AutoML tasks. The results verify the effectiveness of the proposed method.
Tasks	AutoML
Published	2019-05-31
URL	https://arxiv.org/abs/1905.13703v1
PDF	https://arxiv.org/pdf/1905.13703v1.pdf
PWC	https://paperswithcode.com/paper/cascaded-algorithm-selection-and-hyper
Repo
Framework

Balanced One-shot Neural Architecture Optimization


Title	Balanced One-shot Neural Architecture Optimization
Authors	Renqian Luo, Tao Qin, Enhong Chen
Abstract	The ability to rank candidate architectures is the key to the performance of neural architecture search~(NAS). One-shot NAS is proposed to reduce the expense but shows inferior performance against conventional NAS and is not adequately stable. We investigate into this and find that the ranking correlation between architectures under one-shot training and the ones under stand-alone full training is poor, which misleads the algorithm to discover better architectures. Further, we show that the training of architectures of different sizes under the current one-shot method is imbalanced, which causes the evaluated performances of the architectures to be less predictable of their ground-truth performances and affects the ranking correlation heavily. Consequently, we propose Balanced NAO where we introduce balanced training of the supernet during the search procedure to encourage more updates for large architectures than small architectures by sampling architectures in proportion to their model sizes. Comprehensive experiments verify that our proposed method is effective and robust which leads to a more stable search. The final discovered architecture shows significant improvements against baselines with a test error rate of 2.60% on CIFAR-10 and top-1 accuracy of 74.4% on ImageNet under the mobile setting. Code and model checkpoints will be publicly available. The code is available at github.com/renqianluo/NAO_pytorch.
Tasks	Neural Architecture Search
Published	2019-09-24
URL	https://arxiv.org/abs/1909.10815v2
PDF	https://arxiv.org/pdf/1909.10815v2.pdf
PWC	https://paperswithcode.com/paper/understanding-and-improving-one-shot-neural
Repo
Framework