Paper Group ANR 441
Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter. Feasibility of Principal Component Analysis in hand gesture recognition system. Comparative analysis of criteria for filtering time series of word usage frequencies. On the approximation by single hidden layer feedforward neural networks with fixed weights. Regula …
Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter
| Title | Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter |
| Authors | Zeyuan Allen-Zhu |
| Abstract | Given a nonconvex function that is an average of $n$ smooth functions, we design stochastic first-order methods to find its approximate stationary points. The convergence of our new methods depends on the smallest (negative) eigenvalue $-\sigma$ of the Hessian, a parameter that describes how nonconvex the function is. Our methods outperform known results for a range of parameter $\sigma$, and can be used to find approximate local minima. Our result implies an interesting dichotomy: there exists a threshold $\sigma_0$ so that the currently fastest methods for $\sigma>\sigma_0$ and for $\sigma<\sigma_0$ have different behaviors: the former scales with $n^{2/3}$ and the latter scales with $n^{3/4}$. |
| Tasks | Stochastic Optimization |
| Published | 2017-02-02 |
| URL | http://arxiv.org/abs/1702.00763v5 |
| http://arxiv.org/pdf/1702.00763v5.pdf | |
| PWC | https://paperswithcode.com/paper/natasha-faster-non-convex-stochastic |
| Repo | |
| Framework | |
Feasibility of Principal Component Analysis in hand gesture recognition system
| Title | Feasibility of Principal Component Analysis in hand gesture recognition system |
| Authors | Tanu Srivastava, Raj Shree Singh, Sunil Kumar, Pavan Chakraborty |
| Abstract | Nowadays actions are increasingly being handled in electronic ways, instead of physical interaction. From earlier times biometrics is used in the authentication of a person. It recognizes a person by using a human trait associated with it like eyes (by calculating the distance between the eyes) and using hand gestures, fingerprint detection, face detection etc. Advantages of using these traits for identification are that they uniquely identify a person and cannot be forgotten or lost. These are unique features of a human being which are being used widely to make the human life simpler. Hand gesture recognition system is a powerful tool that supports efficient interaction between the user and the computer. The main moto of hand gesture recognition research is to create a system which can recognise specific hand gestures and use them to convey useful information for device control. This paper presents an experimental study over the feasibility of principal component analysis in hand gesture recognition system. PCA is a powerful tool for analyzing data. The primary goal of PCA is dimensionality reduction. Frames are extracted from the Sheffield KInect Gesture (SKIG) dataset. The implementation is done by creating a training set and then training the recognizer. It uses Eigen space by processing the eigenvalues and eigenvectors of the images in training set. Euclidean distance with the threshold value is used as similarity metric to recognize the gestures. The experimental results show that PCA is feasible to be used for hand gesture recognition system. |
| Tasks | Dimensionality Reduction, Face Detection, Gesture Recognition, Hand Gesture Recognition, Hand-Gesture Recognition |
| Published | 2017-02-23 |
| URL | http://arxiv.org/abs/1702.07371v1 |
| http://arxiv.org/pdf/1702.07371v1.pdf | |
| PWC | https://paperswithcode.com/paper/feasibility-of-principal-component-analysis |
| Repo | |
| Framework | |
Comparative analysis of criteria for filtering time series of word usage frequencies
| Title | Comparative analysis of criteria for filtering time series of word usage frequencies |
| Authors | Inna A. Belashova, Vladimir V. Bochkarev |
| Abstract | This paper describes a method of nonlinear wavelet thresholding of time series. The Ramachandran-Ranganathan runs test is used to assess the quality of approximation. To minimize the objective function, it is proposed to use genetic algorithms - one of the stochastic optimization methods. The suggested method is tested both on the model series and on the word frequency series using the Google Books Ngram data. It is shown that method of filtering which uses the runs criterion shows significantly better results compared with the standard wavelet thresholding. The method can be used when quality of filtering is of primary importance but not the speed of calculations. |
| Tasks | Stochastic Optimization, Time Series |
| Published | 2017-12-10 |
| URL | http://arxiv.org/abs/1712.03512v1 |
| http://arxiv.org/pdf/1712.03512v1.pdf | |
| PWC | https://paperswithcode.com/paper/comparative-analysis-of-criteria-for |
| Repo | |
| Framework | |
On the approximation by single hidden layer feedforward neural networks with fixed weights
| Title | On the approximation by single hidden layer feedforward neural networks with fixed weights |
| Authors | Namig J. Guliyev, Vugar E. Ismailov |
| Abstract | Feedforward neural networks have wide applicability in various disciplines of science due to their universal approximation property. Some authors have shown that single hidden layer feedforward neural networks (SLFNs) with fixed weights still possess the universal approximation property provided that approximated functions are univariate. But this phenomenon does not lay any restrictions on the number of neurons in the hidden layer. The more this number, the more the probability of the considered network to give precise results. In this note, we constructively prove that SLFNs with the fixed weight $1$ and two neurons in the hidden layer can approximate any continuous function on a compact subset of the real line. The applicability of this result is demonstrated in various numerical examples. Finally, we show that SLFNs with fixed weights cannot approximate all continuous multivariate functions. |
| Tasks | |
| Published | 2017-08-21 |
| URL | http://arxiv.org/abs/1708.06219v1 |
| http://arxiv.org/pdf/1708.06219v1.pdf | |
| PWC | https://paperswithcode.com/paper/on-the-approximation-by-single-hidden-layer |
| Repo | |
| Framework | |
Regularization for Deep Learning: A Taxonomy
| Title | Regularization for Deep Learning: A Taxonomy |
| Authors | Jan Kukačka, Vladimir Golkov, Daniel Cremers |
| Abstract | Regularization is one of the crucial ingredients of deep learning, yet the term regularization has various definitions, and regularization methods are often studied separately from each other. In our work we present a systematic, unifying taxonomy to categorize existing methods. We distinguish methods that affect data, network architectures, error terms, regularization terms, and optimization procedures. We do not provide all details about the listed methods; instead, we present an overview of how the methods can be sorted into meaningful categories and sub-categories. This helps revealing links and fundamental similarities between them. Finally, we include practical recommendations both for users and for developers of new regularization methods. |
| Tasks | |
| Published | 2017-10-29 |
| URL | http://arxiv.org/abs/1710.10686v1 |
| http://arxiv.org/pdf/1710.10686v1.pdf | |
| PWC | https://paperswithcode.com/paper/regularization-for-deep-learning-a-taxonomy |
| Repo | |
| Framework | |
Depth Assisted Full Resolution Network for Single Image-based View Synthesis
| Title | Depth Assisted Full Resolution Network for Single Image-based View Synthesis |
| Authors | Xiaodong Cun, Feng Xu, Chi-Man Pun, Hao Gao |
| Abstract | Researches in novel viewpoint synthesis majorly focus on interpolation from multi-view input images. In this paper, we focus on a more challenging and ill-posed problem that is to synthesize novel viewpoints from one single input image. To achieve this goal, we propose a novel deep learning-based technique. We design a full resolution network that extracts local image features with the same resolution of the input, which contributes to derive high resolution and prevent blurry artifacts in the final synthesized images. We also involve a pre-trained depth estimation network into our system, and thus 3D information is able to be utilized to infer the flow field between the input and the target image. Since the depth network is trained by depth order information between arbitrary pairs of points in the scene, global image features are also involved into our system. Finally, a synthesis layer is used to not only warp the observed pixels to the desired positions but also hallucinate the missing pixels with recorded pixels. Experiments show that our technique performs well on images of various scenes, and outperforms the state-of-the-art techniques. |
| Tasks | Depth Estimation |
| Published | 2017-11-17 |
| URL | http://arxiv.org/abs/1711.06620v1 |
| http://arxiv.org/pdf/1711.06620v1.pdf | |
| PWC | https://paperswithcode.com/paper/depth-assisted-full-resolution-network-for |
| Repo | |
| Framework | |
Neural Machine Translation with Gumbel-Greedy Decoding
| Title | Neural Machine Translation with Gumbel-Greedy Decoding |
| Authors | Jiatao Gu, Daniel Jiwoong Im, Victor O. K. Li |
| Abstract | Previous neural machine translation models used some heuristic search algorithms (e.g., beam search) in order to avoid solving the maximum a posteriori problem over translation sentences at test time. In this paper, we propose the Gumbel-Greedy Decoding which trains a generative network to predict translation under a trained model. We solve such a problem using the Gumbel-Softmax reparameterization, which makes our generative network differentiable and trainable through standard stochastic gradient methods. We empirically demonstrate that our proposed model is effective for generating sequences of discrete words. |
| Tasks | Machine Translation |
| Published | 2017-06-22 |
| URL | http://arxiv.org/abs/1706.07518v1 |
| http://arxiv.org/pdf/1706.07518v1.pdf | |
| PWC | https://paperswithcode.com/paper/neural-machine-translation-with-gumbel-greedy |
| Repo | |
| Framework | |
On the Integration of Optical Flow and Action Recognition
| Title | On the Integration of Optical Flow and Action Recognition |
| Authors | Laura Sevilla-Lara, Yiyi Liao, Fatma Guney, Varun Jampani, Andreas Geiger, Michael J. Black |
| Abstract | Most of the top performing action recognition methods use optical flow as a “black box” input. Here we take a deeper look at the combination of flow and action recognition, and investigate why optical flow is helpful, what makes a flow method good for action recognition, and how we can make it better. In particular, we investigate the impact of different flow algorithms and input transformations to better understand how these affect a state-of-the-art action recognition method. Furthermore, we fine tune two neural-network flow methods end-to-end on the most widely used action recognition dataset (UCF101). Based on these experiments, we make the following five observations: 1) optical flow is useful for action recognition because it is invariant to appearance, 2) optical flow methods are optimized to minimize end-point-error (EPE), but the EPE of current methods is not well correlated with action recognition performance, 3) for the flow methods tested, accuracy at boundaries and at small displacements is most correlated with action recognition performance, 4) training optical flow to minimize classification error instead of minimizing EPE improves recognition performance, and 5) optical flow learned for the task of action recognition differs from traditional optical flow especially inside the human body and at the boundary of the body. These observations may encourage optical flow researchers to look beyond EPE as a goal and guide action recognition researchers to seek better motion cues, leading to a tighter integration of the optical flow and action recognition communities. |
| Tasks | Optical Flow Estimation, Temporal Action Localization |
| Published | 2017-12-22 |
| URL | http://arxiv.org/abs/1712.08416v1 |
| http://arxiv.org/pdf/1712.08416v1.pdf | |
| PWC | https://paperswithcode.com/paper/on-the-integration-of-optical-flow-and-action |
| Repo | |
| Framework | |
Application of Fuzzy Assessing for Reliability Decision Making
| Title | Application of Fuzzy Assessing for Reliability Decision Making |
| Authors | Shoele Jamali, Mehrdad J. Bani |
| Abstract | This paper proposes a new fuzzy assessing procedure with application in management decision making. The proposed fuzzy approach build the membership functions for system characteristics of a standby repairable system. This method is used to extract a family of conventional crisp intervals from the fuzzy repairable system for the desired system characteristics. This can be determined with a set of nonlinear parametric programing using the membership functions. When system characteristics are governed by the membership functions, more information is provided for use by management, and because the redundant system is extended to the fuzzy environment, general repairable systems are represented more accurately and the analytic results are more useful for designers and practitioners. Also beside standby, active redundancy systems are used in many cases so this article has many practical instances. Different from other studies, our model provides, a good estimated value based on uncertain environments, a comparison discussion of using fuzzy theory and conventional method and also a comparison between parallel (active redundancy) and series system in fuzzy world when we have standby redundancy. When the membership function intervals cannot be inverted explicitly, system management or designers can specify the system characteristics of interest, perform numerical calculations, examine the corresponding {\alpha}-cuts, and use this information to develop or improve system processes. |
| Tasks | Decision Making |
| Published | 2017-07-06 |
| URL | http://arxiv.org/abs/1707.01727v1 |
| http://arxiv.org/pdf/1707.01727v1.pdf | |
| PWC | https://paperswithcode.com/paper/application-of-fuzzy-assessing-for |
| Repo | |
| Framework | |
Scaled Nuclear Norm Minimization for Low-Rank Tensor Completion
| Title | Scaled Nuclear Norm Minimization for Low-Rank Tensor Completion |
| Authors | Morteza Ashraphijuo, Xiaodong Wang |
| Abstract | Minimizing the nuclear norm of a matrix has been shown to be very efficient in reconstructing a low-rank sampled matrix. Furthermore, minimizing the sum of nuclear norms of matricizations of a tensor has been shown to be very efficient in recovering a low-Tucker-rank sampled tensor. In this paper, we propose to recover a low-TT-rank sampled tensor by minimizing a weighted sum of nuclear norms of unfoldings of the tensor. We provide numerical results to show that our proposed method requires significantly less number of samples to recover to the original tensor in comparison with simply minimizing the sum of nuclear norms since the structure of the unfoldings in the TT tensor model is fundamentally different from that of matricizations in the Tucker tensor model. |
| Tasks | |
| Published | 2017-07-25 |
| URL | http://arxiv.org/abs/1707.07976v1 |
| http://arxiv.org/pdf/1707.07976v1.pdf | |
| PWC | https://paperswithcode.com/paper/scaled-nuclear-norm-minimization-for-low-rank |
| Repo | |
| Framework | |
Saliency Weighted Convolutional Features for Instance Search
| Title | Saliency Weighted Convolutional Features for Instance Search |
| Authors | Eva Mohedano, Kevin McGuinness, Xavier Giro-i-Nieto, Noel E. O’Connor |
| Abstract | This work explores attention models to weight the contribution of local convolutional representations for the instance search task. We present a retrieval framework based on bags of local convolutional features (BLCF) that benefits from saliency weighting to build an efficient image representation. The use of human visual attention models (saliency) allows significant improvements in retrieval performance without the need to conduct region analysis or spatial verification, and without requiring any feature fine tuning. We investigate the impact of different saliency models, finding that higher performance on saliency benchmarks does not necessarily equate to improved performance when used in instance search tasks. The proposed approach outperforms the state-of-the-art on the challenging INSTRE benchmark by a large margin, and provides similar performance on the Oxford and Paris benchmarks compared to more complex methods that use off-the-shelf representations. The source code used in this project is available at https://imatge-upc.github.io/salbow/ |
| Tasks | Instance Search |
| Published | 2017-11-29 |
| URL | http://arxiv.org/abs/1711.10795v1 |
| http://arxiv.org/pdf/1711.10795v1.pdf | |
| PWC | https://paperswithcode.com/paper/saliency-weighted-convolutional-features-for |
| Repo | |
| Framework | |
Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning
| Title | Wild Patterns: Ten Years After the Rise of Adversarial Machine Learning |
| Authors | Battista Biggio, Fabio Roli |
| Abstract | Learning-based pattern classifiers, including deep networks, have shown impressive performance in several application domains, ranging from computer vision to cybersecurity. However, it has also been shown that adversarial input perturbations carefully crafted either at training or at test time can easily subvert their predictions. The vulnerability of machine learning to such wild patterns (also referred to as adversarial examples), along with the design of suitable countermeasures, have been investigated in the research field of adversarial machine learning. In this work, we provide a thorough overview of the evolution of this research area over the last ten years and beyond, starting from pioneering, earlier work on the security of non-deep learning algorithms up to more recent work aimed to understand the security properties of deep learning algorithms, in the context of computer vision and cybersecurity tasks. We report interesting connections between these apparently-different lines of work, highlighting common misconceptions related to the security evaluation of machine-learning algorithms. We review the main threat models and attacks defined to this end, and discuss the main limitations of current work, along with the corresponding future challenges towards the design of more secure learning algorithms. |
| Tasks | |
| Published | 2017-12-08 |
| URL | http://arxiv.org/abs/1712.03141v2 |
| http://arxiv.org/pdf/1712.03141v2.pdf | |
| PWC | https://paperswithcode.com/paper/wild-patterns-ten-years-after-the-rise-of |
| Repo | |
| Framework | |
Learning Compositional Visual Concepts with Mutual Consistency
| Title | Learning Compositional Visual Concepts with Mutual Consistency |
| Authors | Yunye Gong, Srikrishna Karanam, Ziyan Wu, Kuan-Chuan Peng, Jan Ernst, Peter C. Doerschuk |
| Abstract | Compositionality of semantic concepts in image synthesis and analysis is appealing as it can help in decomposing known and generatively recomposing unknown data. For instance, we may learn concepts of changing illumination, geometry or albedo of a scene, and try to recombine them to generate physically meaningful, but unseen data for training and testing. In practice however we often do not have samples from the joint concept space available: We may have data on illumination change in one data set and on geometric change in another one without complete overlap. We pose the following question: How can we learn two or more concepts jointly from different data sets with mutual consistency where we do not have samples from the full joint space? We present a novel answer in this paper based on cyclic consistency over multiple concepts, represented individually by generative adversarial networks (GANs). Our method, ConceptGAN, can be understood as a drop in for data augmentation to improve resilience for real world applications. Qualitative and quantitative evaluations demonstrate its efficacy in generating semantically meaningful images, as well as one shot face verification as an example application. |
| Tasks | Data Augmentation, Face Verification, Image Generation |
| Published | 2017-11-16 |
| URL | http://arxiv.org/abs/1711.06148v2 |
| http://arxiv.org/pdf/1711.06148v2.pdf | |
| PWC | https://paperswithcode.com/paper/learning-compositional-visual-concepts-with |
| Repo | |
| Framework | |
Unsupervised learning of object landmarks by factorized spatial embeddings
| Title | Unsupervised learning of object landmarks by factorized spatial embeddings |
| Authors | James Thewlis, Hakan Bilen, Andrea Vedaldi |
| Abstract | Learning automatically the structure of object categories remains an important open problem in computer vision. In this paper, we propose a novel unsupervised approach that can discover and learn landmarks in object categories, thus characterizing their structure. Our approach is based on factorizing image deformations, as induced by a viewpoint change or an object deformation, by learning a deep neural network that detects landmarks consistently with such visual effects. Furthermore, we show that the learned landmarks establish meaningful correspondences between different object instances in a category without having to impose this requirement explicitly. We assess the method qualitatively on a variety of object types, natural and man-made. We also show that our unsupervised landmarks are highly predictive of manually-annotated landmarks in face benchmark datasets, and can be used to regress these with a high degree of accuracy. |
| Tasks | Unsupervised Facial Landmark Detection |
| Published | 2017-05-05 |
| URL | http://arxiv.org/abs/1705.02193v2 |
| http://arxiv.org/pdf/1705.02193v2.pdf | |
| PWC | https://paperswithcode.com/paper/unsupervised-learning-of-object-landmarks-by |
| Repo | |
| Framework | |
Identifying hazardousness of sewer pipeline gas mixture using classification methods: a comparative study
| Title | Identifying hazardousness of sewer pipeline gas mixture using classification methods: a comparative study |
| Authors | Varun Kumar Ojha, Paramartha Dutta, Atal Chaudhuri |
| Abstract | In this work, we formulated a real-world problem related to sewer pipeline gas detection using the classification-based approaches. The primary goal of this work was to identify the hazardousness of sewer pipeline to offer safe and non-hazardous access to sewer pipeline workers so that the human fatalities, which occurs due to the toxic exposure of sewer gas components, can be avoided. The dataset acquired through laboratory tests, experiments, and various literature sources was organized to design a predictive model that was able to identify/classify hazardous and non-hazardous situation of sewer pipeline. To design such prediction model, several classification algorithms were used and their performances were evaluated and compared, both empirically and statistically, over the collected dataset. In addition, the performances of several ensemble methods were analyzed to understand the extent of improvement offered by these methods. The result of this comprehensive study showed that the instance-based learning algorithm performed better than many other algorithms such as multilayer perceptron, radial basis function network, support vector machine, reduced pruning tree. Similarly, it was observed that multi-scheme ensemble approach enhanced the performance of base predictors. |
| Tasks | |
| Published | 2017-05-16 |
| URL | http://arxiv.org/abs/1707.00561v1 |
| http://arxiv.org/pdf/1707.00561v1.pdf | |
| PWC | https://paperswithcode.com/paper/identifying-hazardousness-of-sewer-pipeline |
| Repo | |
| Framework | |
