results for au:Yang_X in:cs

- Mar 20 2018 cs.HC arXiv:1803.06625v1Traditional instrument learning is time-consuming. It begins with learning music notation and necessitates layers of sophistication and abstraction. Haptic interfaces open another door to the music world for the vast majority of beginners when traditional training methods are not effective. However, existing haptic interfaces can only deal with specially designed pieces with great restrictions on performance duration and pitch range due to the fact that not all performance motions could be guided haptically for most instruments. Our system breaks such restrictions using a semi-haptic interface. For the first time, the pitch range of the haptically learned pieces goes beyond an octave (with the fingering motion covers most of the possible choices) and the duration of learned pieces cover a whole phrase. This significant change leads to a more realistic instrument learning process. Experiments show that our semi-haptic interface is effective as long as learners are not "tone deaf." Using our prototype device, the learning rate is about 30% faster compared to learning from videos.
- Mar 20 2018 cs.CV arXiv:1803.06798v1This paper studies the object transfiguration problem in wild images. The generative network in classical GANs for object transfiguration often undertakes a dual responsibility: to detect the objects of interests and to convert the object from source domain to target domain. In contrast, we decompose the generative network into two separat networks, each of which is only dedicated to one particular sub-task. The attention network predicts spatial attention maps of images, and the transformation network focuses on translating objects. Attention maps produced by attention network are encouraged to be sparse, so that major attention can be paid to objects of interests. No matter before or after object transfiguration, attention maps should remain constant. In addition, learning attention network can receive more instructions, given the available segmentation annotations of images. Experimental results demonstrate the necessity of investigating attention in object transfiguration, and that the proposed algorithm can learn accurate attention to improve quality of generated images.
- Mar 14 2018 cs.CV arXiv:1803.04793v1Sparse representation classification achieves good results by addressing recognition problem with sufficient training samples per subject. However, SRC performs not very well for small sample data. In this paper, an inverse-projection group sparse representation model is presented for breast tumor classification, which is based on constructing low-rank variation dictionary. The proposed low-rank variation dictionary tackles tumor recognition problem from the viewpoint of detecting and using variations in gene expression profiles of normal and patients, rather than directly using these samples. The inverse projection group sparsity representation model is constructed based on taking full using of exist samples and group effect of microarray gene data. Extensive experiments on public breast tumor microarray gene expression datasets demonstrate the proposed technique is competitive with state-of-the-art methods. The results of Breast-1, Breast-2 and Breast-3 databases are 80.81%, 89.10% and 100% respectively, which are better than the latest literature.
- Mar 12 2018 cs.CV arXiv:1803.03562v1Microarray gene expression data-based breast tumor classification is an active and challenging issue. In this paper, a robust breast tumor recognition framework is presented based on considering reducing clinical misdiagnosis rate and exploiting available information in existing samples. A wrapper gene selection method is established from a new perspective of reducing clinical misdiagnosis rate. The further feature selection of information genes is fulfilled by a modified NMF model, which is motivated by hierarchical learning and layer-wise pre-training strategy in deep learning. For completing the classification, an inverse projection sparse representation (IPSR) model is constructed to exploit information embedded in existing samples, especially in test ones. Moreover, the IPSR model is optimized via generalized ADMM and the corresponding convergence is analyzed. Extensive experiments on three public breast tumor datasets show that the proposed method is stable and effective for breast tumor classification. Compared to the latest literature, there is 14% higher in classification accuracy. Specificity and sensitivity achieve 94.17% and 97.5%.
- Mar 02 2018 cs.PL arXiv:1803.00403v2In recent years, a number of lightweight programs have been deployed in critical domains, such as in smart contracts based on blockchain technology. Therefore, the security and reliability of such programs should be guaranteed by the most credible technology. Higher-order logic theorem proving is one of the most reliable technologies for verifying the properties of programs. However, programs may be developed by different high-level programming languages, and a general, extensible, and reusable formal memory (GERM) framework that can simultaneously support different formal verification specifications, particularly at the code level, is presently unavailable for verifying the properties of programs. Therefore, the present work proposes a GERM framework to fill this gap. The framework simulates physical memory hardware structure, including a low-level formal memory space, and provides a set of simple, nonintrusive application programming interfaces and assistant tools using Coq that can support different formal verification specifications simultaneously. The proposed GERM framework is independent and customizable, and was verified entirely in Coq. We also present an extension of Curry-Howard isomorphism, denoted as execution-verification isomorphism (EVI), which combines symbolic execution and theorem proving for increasing the degree of automation in higher-order logic theorem proving assistant tools. We also implement a toy functional programming language in a generalized algebraic datatypes style and a formal interpreter in Coq based on the GERM framework. These implementations are then employed to demonstrate the application of EVI to a simple code segment.
- Feb 26 2018 cs.AI arXiv:1802.08365v2Real-time bidding (RTB) is almost the most important mechanism in online display advertising, where proper bid for each page view plays a vital and essential role for good marketing results. Budget constrained bidding is a typical scenario in RTB mechanism where the advertisers hope to maximize total value of winning impressions under a pre-set budget constraint. However, the optimal strategy is hard to be derived due to complexity and volatility of the auction environment. To address the challenges, in this paper, we formulate budget constrained bidding as a Markov Decision Process. Quite different from prior model-based work, we propose a novel framework based on model-free reinforcement learning which sequentially regulates the bidding parameter rather than directly producing bid. Along this line, we further innovate a reward function which deploys a deep neural network to learn appropriate reward and thus leads the agent to deliver the optimal policy effectively; we also design an adaptive $\epsilon$-greedy strategy which adjusts the exploration behaviour dynamically and further improves the performance. Experimental results on real dataset demonstrate the effectiveness of our framework.
- Many problems in areas as diverse as recommendation systems, social network analysis, semantic search, and distributed root cause analysis can be modeled as pattern search on labeled graphs (also called "heterogeneous information networks" or HINs). Given a large graph and a query pattern with node and edge label constraints, a fundamental challenge is to nd the top-k matches ac- cording to a ranking function over edge and node weights. For users, it is di cult to select value k . We therefore propose the novel notion of an any-k ranking algorithm: for a given time budget, re- turn as many of the top-ranked results as possible. Then, given additional time, produce the next lower-ranked results quickly as well. It can be stopped anytime, but may have to continues until all results are returned. This paper focuses on acyclic patterns over arbitrary labeled graphs. We are interested in practical algorithms that effectively exploit (1) properties of heterogeneous networks, in particular selective constraints on labels, and (2) that the users often explore only a fraction of the top-ranked results. Our solution, KARPET, carefully integrates aggressive pruning that leverages the acyclic nature of the query, and incremental guided search. It enables us to prove strong non-trivial time and space guarantees, which is generally considered very hard for this type of graph search problem. Through experimental studies we show that KARPET achieves running times in the order of milliseconds for tree patterns on large networks with millions of nodes and edges.
- Influence maximization which asks for $k$-size seed set from a social network such that maximizing the influence over all other users (called influence spread) has widely attracted attention due to its significant applications in viral marketing and rumor control. In real world scenarios, people are interested in the most influential users in particular topics, and want to subscribe the topics-of-interests over social networks. In this paper, we formulate the problem of influential users subscription on time-decaying social stream, which asks for maintaining the $k$-size influential users sets for each topic-aware subscription queries. We first analyze the widely adopted sliding window model and propose a newly time-decaying influence model to overcome the shortages when calculating the influence over social stream. Developed from sieve based streaming algorithm, we propose an efficient algorithm to support the calculation of time-decaying influence over dynamically updating social networks. Using information among subscriptions, we then construct the Prefix Tree Structure to allow us minimizing the times of calculating influence of each update and easily maintained. Pruning techniques are also applied to the Prefix Tree to optimize the performance of social stream update. Our approach ensures a $\frac{1}{2}-\epsilon$ approximation ratio. Experimental results show that our approach significantly outperforms the baseline approaches in efficiency and result quality.
- Multi-channel speech enhancement with ad-hoc sensors has been a challenging task. Speech model guided beamforming algorithms are able to recover natural sounding speech, but the speech models tend to be oversimplified or the inference would otherwise be too complicated. On the other hand, deep learning based enhancement approaches are able to learn complicated speech distributions and perform efficient inference, but they are unable to deal with variable number of input channels. Also, deep learning approaches introduce a lot of errors, particularly in the presence of unseen noise types and settings. We have therefore proposed an enhancement framework called DEEPBEAM, which combines the two complementary classes of algorithms. DEEPBEAM introduces a beamforming filter to produce natural sounding speech, but the filter coefficients are determined with the help of a monaural speech enhancement neural network. Experiments on synthetic and real-world data show that DEEPBEAM is able to produce clean, dry and natural sounding speech, and is robust against unseen noise.
- Feb 13 2018 cs.LG arXiv:1802.03452v1The performance of distance-based classifiers heavily depends on the underlying distance metric, so it is valuable to learn a suitable metric from the data. To address the problem of multimodality, it is desirable to learn local metrics. In this short paper, we define a new intuitive distance with local metrics and influential regions, and subsequently propose a novel local metric learning method for distance-based classification. Our key intuition is to partition the metric space into influential regions and a background region, and then regulate the effectiveness of each local metric to be within the related influential regions. We learn local metrics and influential regions to reduce the empirical hinge loss, and regularize the parameters on the basis of a resultant learning bound. Encouraging experimental results are obtained from various public and popular data sets.
- Feb 13 2018 cs.LG arXiv:1802.03464v1In this paper, we propose the Lipschitz margin ratio and a new metric learning framework for classification through maximizing the ratio. This framework enables the integration of both the inter-class margin and the intra-class dispersion, as well as the enhancement of the generalization ability of a classifier. To introduce the Lipschitz margin ratio and its associated learning bound, we elaborate the relationship between metric learning and Lipschitz functions, as well as the representability and learnability of the Lipschitz functions. After proposing the new metric learning framework based on the introduced Lipschitz margin ratio, we also prove that some well known metric learning algorithms can be shown as special cases of the proposed framework. In addition, we illustrate the framework by implementing it for learning the squared Mahalanobis metric, and by demonstrating its encouraging results on eight popular datasets of machine learning.
- The performance of automatic speech recognition systems degrades with increasing mismatch between the training and testing scenarios. Differences in speaker accents are a significant source of such mismatch. The traditional approach to deal with multiple accents involves pooling data from several accents during training and building a single model in multi-task fashion, where tasks correspond to individual accents. In this paper, we explore an alternate model where we jointly learn an accent classifier and a multi-task acoustic model. Experiments on the American English Wall Street Journal and British English Cambridge corpora demonstrate that our joint model outperforms the strong multi-task acoustic model baseline. We obtain a 5.94% relative improvement in word error rate on British English, and 9.47% relative improvement on American English. This illustrates that jointly modeling with accent information improves acoustic model performance.
- Feb 06 2018 cs.PL arXiv:1802.01220v2Most existed work require knowledge about the effect of program instructions (or statements) to analyze and verify algorithms. In this paper, by revealing some findings on executions of object programs, we define two basic concepts -- effect equivalence relation and effect race relation. Further, we show three effect theorems about the race and histories. The core result is that the effect race relation is the accurate relation to capture the internal steps, of which precedence orders are the reason to cause chaotic histories. In addition, the concept -- linearization points -- widely used in the object verification, is defined formally as the typical effect race relation. These results provide a clear basis for analyzing intricate fine-grained executions. We conduct a lot of experiments on real object algorithms to show the accuracy and efficiency of these definitions in practice. A simple quantitative analysis method for these algorithms is also proposed.
- This paper introduces the factorial marked temporal point process model and presents efficient learning methods. In conventional (multi-dimensional) marked temporal point process models, event is often encoded by a single discrete variable i.e. a marker. In this paper, we describe the factorial marked point processes whereby time-stamped event is factored into multiple markers. Accordingly the size of the infectivity matrix modeling the effect between pairwise markers is in power order w.r.t. the number of the discrete marker space. We propose a decoupled learning method with two learning procedures: i) directly solving the model based on two techniques: Alternating Direction Method of Multipliers and Fast Iterative Shrinkage-Thresholding Algorithm; ii) involving a reformulation that transforms the original problem into a Logistic Regression model for more efficient learning. Moreover, a sparse group regularizer is added to identify the key profile features and event labels. Empirical results on real world datasets demonstrate the efficiency of our decoupled and reformulated method. The source code is available online.
- Jan 09 2018 cs.CV arXiv:1801.02021v1Recently, deep learning has achieved very promising results in visual object tracking. Deep neural networks in existing tracking methods require a lot of training data to learn a large number of parameters. However, training data is not sufficient for visual object tracking as annotations of a target object are only available in the first frame of a test sequence. In this paper, we propose to learn hierarchical features for visual object tracking by using tree structure based Recursive Neural Networks (RNN), which have fewer parameters than other deep neural networks, e.g. Convolutional Neural Networks (CNN). First, we learn RNN parameters to discriminate between the target object and background in the first frame of a test sequence. Tree structure over local patches of an exemplar region is randomly generated by using a bottom-up greedy search strategy. Given the learned RNN parameters, we create two dictionaries regarding target regions and corresponding local patches based on the learned hierarchical features from both top and leaf nodes of multiple random trees. In each of the subsequent frames, we conduct sparse dictionary coding on all candidates to select the best candidate as the new target location. In addition, we online update two dictionaries to handle appearance changes of target objects. Experimental results demonstrate that our feature learning algorithm can significantly improve tracking performance on benchmark datasets.
- Daily engagement in life experiences is increasingly interwoven with mobile device use. Screen capture at the scale of seconds is being used in behavioral studies and to implement "just-in-time" health interventions. The increasing psychological breadth of digital information will continue to make the actual screens that people view a preferred if not required source of data about life experiences. Effective and efficient Information Extraction and Retrieval from digital screenshots is a crucial prerequisite to successful use of screen data. In this paper, we present the experimental workflow we exploited to: (i) pre-process a unique collection of screen captures, (ii) extract unstructured text embedded in the images, (iii) organize image text and metadata based on a structured schema, (iv) index the resulting document collection, and (v) allow for Image Retrieval through a dedicated vertical search engine application. The adopted procedure integrates different open source libraries for traditional image processing, Optical Character Recognition (OCR), and Image Retrieval. Our aim is to assess whether and how state-of-the-art methodologies can be applied to this novel data set. We show how combining OpenCV-based pre-processing modules with a Long short-term memory (LSTM) based release of Tesseract OCR, without ad hoc training, led to a 74% character-level accuracy of the extracted text. Further, we used the processed repository as baseline for a dedicated Image Retrieval system, for the immediate use and application for behavioral and prevention scientists. We discuss issues of Text Information Extraction and Retrieval that are particular to the screenshot image case and suggest important future work.
- Dec 27 2017 cs.RO arXiv:1712.09162v1Aerial surveillance and monitoring demand both real-time and robust motion detection from a moving camera. Most existing techniques for drones involve sending a video data streams back to a ground station with a high-end desktop computer or server. These methods share one major drawback: data transmission is subjected to considerable delay and possible corruption. Onboard computation can not only overcome the data corruption problem but also increase the range of motion. Unfortunately, due to limited weight-bearing capacity, equipping drones with computing hardware of high processing capability is not feasible. Therefore, developing a motion detection system with real-time performance and high accuracy for drones with limited computing power is highly desirable. In this paper, we propose a visual-inertial drone system for real-time motion detection, namely REDBEE, that helps overcome challenges in shooting scenes with strong parallax and dynamic background. REDBEE, which can run on the state-of-the-art commercial low-power application processor (e.g. Snapdragon Flight board used for our prototype drone), achieves real-time performance with high detection accuracy. The REDBEE system overcomes obstacles in shooting scenes with strong parallax through an inertial-aided dual-plane homography estimation; it solves the issues in shooting scenes with dynamic background by distinguishing the moving targets through a probabilistic model based on spatial, temporal, and entropy consistency. The experiments are presented which demonstrate that our system obtains greater accuracy when detecting moving targets in outdoor environments than the state-of-the-art real-time onboard detection systems.
- Dec 27 2017 cs.CL arXiv:1712.08841v2Characters have commonly been regarded as the minimal processing unit in Natural Language Processing (NLP). But many non-latin languages have hieroglyphic writing systems, involving a big alphabet with thousands or millions of characters. Each character is composed of even smaller parts, which are often ignored by the previous work. In this paper, we propose a novel architecture employing two stacked Long Short-Term Memory Networks (LSTMs) to learn sub-character level representation and capture deeper level of semantic meanings. To build a concrete study and substantiate the efficiency of our neural architecture, we take Chinese Word Segmentation as a research case example. Among those languages, Chinese is a typical case, for which every character contains several components called radicals. Our networks employ a shared radical level embedding to solve both Simplified and Traditional Chinese Word Segmentation, without extra Traditional to Simplified Chinese conversion, in such a highly end-to-end way the word segmentation can be significantly simplified compared to the previous work. Radical level embeddings can also capture deeper semantic meaning below character level and improve the system performance of learning. By tying radical and character embeddings together, the parameter count is reduced whereas semantic knowledge is shared and transferred between two levels, boosting the performance largely. On 3 out of 4 Bakeoff 2005 datasets, our method surpassed state-of-the-art results by up to 0.4%. Our results are reproducible, source codes and corpora are available on GitHub.
- Dec 13 2017 cs.NI arXiv:1712.04201v1We investigate network performance of ultra-dense heterogeneous networks (HetNets) and study the maximum energy-efficient base station (BS) deployment incorporating probabilistic non-line-of-sight (NLoS) and line-of-sight (LoS) transmissions. First, we develop an analytical framework with the maximum instantaneous received power (MIRP) and the maximum average received power (MARP) association schemes to model the coverage probability and related performance metrics, e.g., the potential throughput (PT) and the energy efficiency (EE). Second, we formulate two optimization problems to achieve the maximum energy-efficient deployment solution with specific service criteria. Simulation results show that there are tradeoffs among the coverage probability, the total power consumption, and the EE. To be specific, the maximum coverage probability with ideal power consumption is superior to that with practical power consumption when the total power constraint is small and inferior to that with practical power consumption when the total power constraint becomes large. Moreover, the maximum EE is a decreasing function with respect to the coverage probability constraint.
- Dec 04 2017 cs.CV arXiv:1712.00097v1In this paper, we address the challenging problem of effi- cient temporal activity detection in untrimmed long videos. While most recent work has focused and advanced the de- tection accuracy, the inference time can take seconds to minutes in processing one video, which is computationally prohibitive for many applications with tight runtime con- straints. This motivates our proposed budget-aware frame- work, which learns to perform activity detection by intel- ligently selecting a small subset of frames according to a specified time or computational budget. We formulate this problem as a Markov decision process, and adopt a recurrent network to model a policy for the frame selec- tion. To train the network, we employ a recurrent policy gradient approach to approximate the gradient of the non- decomposable and non-differentiable objective defined in our problem. In the extensive experiments on two bench- mark datasets, we achieve competitive detection accuracy, and more importantly, our approach is able to substantially reduce computational time and detect multiple activities in only 348ms for each untrimmed long video of THUMOS14 and ActivityNet.
- Computer poetry generation is our first step towards computer writing. Writing must have a theme. The current approaches of using sequence-to-sequence models with attention often produce non-thematic poems. We present a novel conditional variational autoencoder with a hybrid decoder adding the deconvolutional neural networks to the general recurrent neural networks to fully learn topic information via latent variables. This approach significantly improves the relevance of the generated poems by representing each line of the poem not only in a context-sensitive manner but also in a holistic way that is highly related to the given keyword and the learned topic. A proposed augmented word2vec model further improves the rhythm and symmetry. Tests show that the generated poems by our approach are mostly satisfying with regulated rules and consistent themes, and 73.42% of them receive an Overall score no less than 3 (the highest score is 5).
- Nov 17 2017 cs.CL arXiv:1711.06061v1Machine translation is going through a radical revolution, driven by the explosive development of deep learning techniques using Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN). In this paper, we consider a special case in machine translation problems, targeting to translate natural language into Structural Query Language (SQL) for data retrieval over relational database. Although generic CNN and RNN learn the grammar structure of SQL when trained with sufficient samples, the accuracy and training efficiency of the model could be dramatically improved, when the translation model is deeply integrated with the grammar rules of SQL. We present a new encoder-decoder framework, with a suite of new approaches, including new semantic features fed into the encoder as well as new grammar-aware states injected into the memory of decoder. These techniques help the neural network focus on understanding semantics of the operations in natural language and save the efforts on SQL grammar learning. The empirical evaluation on real world database and queries show that our approach outperform state-of-the-art solution by a significant margin.
- We present a robust and precise localization system that achieves centimeter-level localization accuracy in disparate city scenes. Our system adaptively uses information from complementary sensors such as GNSS, LiDAR, and IMU to achieve high localization accuracy and resilience in challenging scenes, such as urban downtown, highways, and tunnels. Rather than relying only on LiDAR intensity or 3D geometry, we make innovative use of LiDAR intensity and altitude cues to significantly improve localization system accuracy and robustness. Our GNSS RTK module utilizes the help of the multi-sensor fusion framework and achieves a better ambiguity resolution success rate. An error-state Kalman filter is applied to fuse the localization measurements from different sources with novel uncertainty estimation. We validate, in detail, the effectiveness of our approaches, achieving 5-10cm RMS accuracy and outperforming previous state-of-the-art systems. Importantly, our system, while deployed in a large autonomous driving fleet, made our vehicles fully autonomous in crowded city streets despite road construction that occurred from time to time. A dataset including more than 60 km real traffic driving in various urban roads is used to comprehensively test our system.
- Deformable image registration and regression are important tasks in medical image analysis. However, they are computationally expensive, especially when analyzing large-scale datasets that contain thousands of images. Hence, cluster computing is typically used, making the approaches dependent on such computational infrastructure. Even larger computational resources are required as study sizes increase. This limits the use of deformable image registration and regression for clinical applications and as component algorithms for other image analysis approaches. We therefore propose using a fast predictive approach to perform image registrations. In particular, we employ these fast registration predictions to approximate a simplified geodesic regression model to capture longitudinal brain changes. The resulting method is orders of magnitude faster than the standard optimization-based regression model and hence facilitates large-scale analysis on a single graphics processing unit (GPU). We evaluate our results on 3D brain magnetic resonance images (MRI) from the ADNI datasets.
- Nov 16 2017 cs.MM arXiv:1711.04916v2The distortion in steganography that usually comes from the modification or recoding on the cover image during the embedding process leaves the steganalyzer with possibility of discriminating. Faced with such a risk, we propose generative steganography with Kerckhoffs' principle (GSK) in this letter. In GSK, the secret messages are generated by a cover image using a generator rather than embedded into the cover, thus resulting in no modifications in the cover. To ensure the security, the generators are trained to meet Kerckhoffs' principle based on generative adversarial networks (GAN). Everything about the GSK system, except the extraction key, is public knowledge for the receivers. The secret messages can be outputted by the generator if and only if the extraction key and the cover image are both inputted. In the generator training procedures, there are two GANs, Message- GAN and Cover-GAN, designed to work jointly making the generated results under the control of the extraction key and the cover image. We provide experimental results on the training process and give an example of the working process by adopting a generator trained on MNIST, which demonstrate that GSK can use a cover image without any modification to generate messages, and without the extraction key or the cover image, only meaningless results would be obtained.
- Most mainstream Automatic Speech Recognition (ASR) systems consider all feature frames equally important. However, acoustic landmark theory is based on a contradictory idea, that some frames are more important than others. Acoustic landmark theory exploits the quantal nonlinear articulatory-acoustic relationships from human speech perception experiments, and provides theoretical support for extracting acoustic features in the vicinity of landmark regions where an abrupt change occurs in the spectrum of speech signals. In this work, we conduct experiments on the TIMIT corpus, with both GMM and DNN based ASR systems and found that frames containing landmarks are more informative than others. We found that altering the level of emphasis on landmarks through accordingly re-weighting acoustic likelihood in frames, tends to reduce the phone error rate (PER). Furthermore, by leveraging the landmark as a heuristic, one of our hybrid DNN frame dropping strategies maintained a PER within 0.44% of optimal when scoring less than half (41.2% to be precise) of the frames. This hybrid strategy out-performs other non-heuristicbased methods and demonstrates the potential of landmarks for reducing computation.
- Oct 03 2017 physics.chem-ph cs.CE arXiv:1710.00616v1Automatic design of organic materials requires black-box optimization in a vast chemical space. In conventional molecular design algorithms, a molecule is built as a combination of predetermined fragments. Recently, deep neural network models such as variational auto encoders (VAEs) and recurrent neural networks (RNNs) are shown to be effective in de novo design of molecules without any predetermined fragments. This paper presents a novel python library ChemTS that explores the chemical space by combining Monte Carlo tree search (MCTS) and an RNN. In a benchmarking problem of optimizing the octanol-water partition coefficient and synthesizability, our algorithm showed superior efficiency in finding high-scoring molecules. ChemTS is available at https://github.com/tsudalab/ChemTS.
- Sep 28 2017 cs.CV arXiv:1709.09550v1We propose the segmentation of noisy datasets into Multiple Inlier Structures with a new Robust Estimator (MISRE). The scale of each individual structure is estimated adaptively from the input data and refined by mean shift, without tuning any parameter in the process, or manually specifying thresholds for different estimation problems. Once all the data points were classified into separate structures, these structures are sorted by their densities with the strongest inlier structures coming out first. Several 2D and 3D synthetic and real examples are presented to illustrate the efficiency, robustness and the limitations of the MISRE algorithm.
- Decentralized control of robots has attracted huge research interests. However, some of the research used unrealistic assumptions without collision avoidance. This report focuses on the collision-free control for multiple robots in both complete coverage and search tasks in 2D and 3D areas which are arbitrary unknown. All algorithms are decentralized as robots have limited abilities and they are mathematically proved. The report starts with the grid selection in the two tasks. Grid patterns simplify the representation of the area and robots only need to move straightly between neighbor vertices. For the 100% complete 2D coverage, the equilateral triangular grid is proposed. For the complete coverage ignoring the boundary effect, the grid with the fewest vertices is calculated in every situation for both 2D and 3D areas. The second part is for the complete coverage in 2D and 3D areas. A decentralized collision-free algorithm with the above selected grid is presented driving robots to sections which are furthest from the reference point. The area can be static or expanding, and the algorithm is simulated in MATLAB. Thirdly, three grid-based decentralized random algorithms with collision avoidance are provided to search targets in 2D or 3D areas. The number of targets can be known or unknown. In the first algorithm, robots choose vacant neighbors randomly with priorities on unvisited ones while the second one adds the repulsive force to disperse robots if they are close. In the third algorithm, if surrounded by visited vertices, the robot will use the breadth-first search algorithm to go to one of the nearest unvisited vertices via the grid. The second search algorithm is verified on Pioneer 3-DX robots. The general way to generate the formula to estimate the search time is demonstrated. Algorithms are compared with five other algorithms in MATLAB to show their effectiveness.
- The new cyber attack pattern of advanced persistent threat (APT) has posed a serious threat to modern society. This paper addresses the APT defense problem, i.e., the problem of how to effectively defend against an APT campaign. Based on a novel APT attack-defense model, the effectiveness of an APT defense strategy is quantified. Thereby, the APT defense problem is modeled as an optimal control problem, in which an optimal control stands for a most effective APT defense strategy. The existence of an optimal control is proved, and an optimality system is derived. Consequently, an optimal control can be figured out by solving the optimality system. Some examples of the optimal control are given. Finally, the influence of some factors on the effectiveness of an optimal control is examined through computer experiments. These findings help organizations to work out policies of defending against APTs.
- Sep 11 2017 cs.SI arXiv:1709.02767v2This paper addresses the issue of suppressing a rumor using the truth in a cost-effective way. First, an individual-level dynamical model capturing the rumor-truth mixed spreading processes is proposed. On this basis, the cost-effective rumor-containing problem is modeled as an optimization problem. Extensive experiments show that finding a cost-effective rumor-containing strategy boils down to enhancing the first truth-spreading rate until the cost effectiveness of the rumor-containing strategy reaches the first turning point. This finding greatly reduces the time spent for solving the optimization problem. The influences of different factors on the optimal cost effectiveness of a rumor-containing strategy are examined through computer simulations. We believe our findings help suppress rumors in a cost-effective way. To our knowledge, this is the first time the rumor-containing problem is treated this way.
- Sep 08 2017 cs.CV arXiv:1709.02371v2We present a compact but effective CNN model for optical flow, called PWC-Net. PWC-Net has been designed according to simple and well-established principles: pyramidal processing, warping, and the use of a cost volume. Cast in a learnable feature pyramid, PWC-Net uses the current optical flow estimate to warp the CNN features of the second image. It then uses the warped features and features of the first image to construct the cost volume, which is processed by a CNN to estimate the optical flow. PWCNet is 17 times smaller in size and easier to train than the recent FlowNet2 model. Moreover, it outperforms all published methods on the MPI Sintel final pass and KITTI 2015 benchmarks, running at about 35 fps on Sintel resolution (1024x436) images. Our model will be publicly available.
- Aug 30 2017 cs.CV arXiv:1708.08687v1Input binarization has shown to be an effective way for network acceleration. However, previous binarization scheme could be regarded as simple pixel-wise thresholding operations (i.e., order-one approximation) and suffers a big accuracy loss. In this paper, we propose a highorder binarization scheme, which achieves more accurate approximation while still possesses the advantage of binary operation. In particular, the proposed scheme recursively performs residual quantization and yields a series of binary input images with decreasing magnitude scales. Accordingly, we propose high-order binary filtering and gradient propagation operations for both forward and backward computations. Theoretical analysis shows approximation error guarantee property of proposed method. Extensive experimental results demonstrate that the proposed scheme yields great recognition accuracy while being accelerated.
- Aug 11 2017 cs.DC arXiv:1708.03184v2Big data analytics on geographically distributed datasets (across data centers or clusters) has been attracting increasing interests from both academia and industry, but also significantly complicates the system and algorithm designs. In this article, we systematically investigate the geo-distributed big-data analytics framework by analyzing the fine-grained paradigm and the key design principles. We present a dynamic global manager selection algorithm (GMSA) to minimize energy consumption cost by fully exploiting the system diversities in geography and variation over time. The algorithm makes real-time decisions based on the measurable system parameters through stochastic optimization methods, while achieving the performance balances between energy cost and latency. Extensive trace-driven simulations verify the effectiveness and efficiency of the proposed algorithm. We also highlight several potential research directions that remain open and require future elaborations in analyzing geo-distributed big data.
- We develop an extension of recently developed methods for obtaining time-space tradeoff lower bounds for problems of learning from random test samples to handle the situation where the space of tests is signficantly smaller than the space of inputs, a class of learning problems that is not handled by prior work. This extension is based on a measure of how matrices amplify the 2-norms of probability distributions that is more refined than the 2-norms of these matrices. As applications that follow from our new technique, we show that any algorithm that learns $m$-variate homogeneous polynomial functions of degree at most $d$ over $\mathbb{F}_2$ from evaluations on randomly chosen inputs either requires space $\Omega(mn)$ or $2^{\Omega(m)}$ time where $n=m^{\Theta(d)}$ is the dimension of the space of such functions. These bounds are asymptotically optimal since they match the tradeoffs achieved by natural learning algorithms for the problems.
- We propose a method for learning Markov network structures for continuous data without invoking any assumptions about the distribution of the variables. The method makes use of previous work on a non-parametric estimator for mutual information which is used to create a non-parametric test for multivariate conditional independence. This independence test is then combined with an efficient constraint-based algorithm for learning the graph structure. The performance of the method is evaluated on several synthetic data sets and it is shown to learn considerably more accurate structures than competing methods when the dependencies between the variables involve non-linearities.
- Aug 03 2017 cs.CV arXiv:1708.00813v1Sustainability of the global environment is dependent on the accurate land cover information over large areas. Even with the increased number of satellite systems and sensors acquiring data with improved spectral, spatial, radiometric and temporal characteristics and the new data distribution policy, most existing land cover datasets were derived from a pixel-based single-date multi-spectral remotely sensed image with low accuracy. To improve the accuracy, the bottleneck is how to develop an accurate and effective image classification technique. By incorporating and utilizing the complete multi-spectral, multi-temporal and spatial information in remote sensing images and considering their inherit spatial and sequential interdependence, we propose a new patch-based RNN (PB-RNN) system tailored for multi-temporal remote sensing data. The system is designed by incorporating distinctive characteristics in multi-temporal remote sensing data. In particular, it uses multi-temporal-spectral-spatial samples and deals with pixels contaminated by clouds/shadow present in the multi-temporal data series. Using a Florida Everglades ecosystem study site covering an area of 771 square kilo-meters, the proposed PB-RNN system has achieved a significant improvement in the classification accuracy over pixel-based RNN system, pixel-based single-imagery NN system, pixel-based multi-images NN system, patch-based single-imagery NN system and patch-based multi-images NN system. For example, the proposed system achieves 97.21% classification accuracy while a pixel-based single-imagery NN system achieves 64.74%. By utilizing methods like the proposed PB-RNN one, we believe that much more accurate land cover datasets can be produced over large areas efficiently.
- Mobile edge computing (MEC) is expected to be an effective solution to deliver 360-degree virtual reality (VR) videos over wireless networks. In contrast to previous computation-constrained MEC framework, which reduces the computation-resource consumption at the mobile VR device by increasing the communication-resource consumption, we develop a communications-constrained MEC framework to reduce communication-resource consumption by increasing the computation-resource consumption and exploiting the caching resources at the mobile VR device in this paper. Specifically, according to the task modularization, the MEC server can only deliver the components which have not been stored in the VR device, and then the VR device uses the received components and the corresponding cached components to construct the task, resulting in low communication-resource consumption but high delay. The MEC server can also compute the task by itself to reduce the delay, however, it consumes more communication-resource due to the delivery of entire task. Therefore, we then propose a task scheduling strategy to decide which computation model should the MEC server operates, in order to minimize the communication-resource consumption under the delay constraint. Finally, we discuss the tradeoffs between communications, computing, and caching in the proposed system.
- Aug 03 2017 cs.CV arXiv:1708.00573v1Automatic and accurate whole-heart and great vessel segmentation from 3D cardiac magnetic resonance (MR) images plays an important role in the computer-assisted diagnosis and treatment of cardiovascular disease. However, this task is very challenging due to ambiguous cardiac borders and large anatomical variations among different subjects. In this paper, we propose a novel densely-connected volumetric convolutional neural network, referred as DenseVoxNet, to automatically segment the cardiac and vascular structures from 3D cardiac MR images. The DenseVoxNet adopts the 3D fully convolutional architecture for effective volume-to-volume prediction. From the learning perspective, our DenseVoxNet has three compelling advantages. First, it preserves the maximum information flow between layers by a densely-connected mechanism and hence eases the network training. Second, it avoids learning redundant feature maps by encouraging feature reuse and hence requires fewer parameters to achieve high performance, which is essential for medical applications with limited training data. Third, we add auxiliary side paths to strengthen the gradient propagation and stabilize the learning process. We demonstrate the effectiveness of DenseVoxNet by comparing it with the state-of-the-art approaches from HVSMR 2016 challenge in conjunction with MICCAI, and our network achieves the best dice coefficient. We also show that our network can achieve better performance than other 3D ConvNets but with fewer parameters.
- It has been an active research issue for many years to construct new bent functions. For $k$ odd with $\gcd(n, k)=1$, and $a\in\mathbb{F}_{3^n}^{*}$, the function $f(x)=Tr(ax^{\frac{3^k+1}{2}})$ is weakly regular bent over $\mathbb{F}_{3^n}$, where $Tr(\cdot):\mathbb{F}_{3^n}\rightarrow\mathbb{F}_3$ is the trace function. This is the well-known Coulter-Matthews bent function. In this paper, we determine the dual function of $f(x)$ completely. As a consequence, we find many classes of ternary bent functions not reported in the literature previously. Such bent functions are not quadratic if $k>1$, and have $\left(\left(\frac{1+\sqrt{5}}{2}\right)^{w+1}-\right.$ $\left.\left(\frac{1-\sqrt{5}}{2}\right)^{w+1}\right)/\sqrt{5}$ or $\left(\left(\frac{1+\sqrt{5}}{2}\right)^{n-w+1}-\right.$ $\left.\left(\frac{1-\sqrt{5}}{2}\right)^{n-w+1}\right)/\sqrt{5}$ trace terms, where $0<w<n$ and $wk\equiv 1\ (\bmod\;n)$. Among them, five special cases are especially interesting: for the case of $k=(n+1)/2$, the number of trace terms is $\left(\left(\frac{1+\sqrt{5}}{2}\right)^{n-1}-\right.$ $\left.\left(\frac{1-\sqrt{5}}{2}\right)^{n-1}\right)/\sqrt{5}$; for the case of $k=n-1$, the number of trace terms is $\left(\left(\frac{1+\sqrt{5}}{2}\right)^n-\right.$ $\left.\left(\frac{1-\sqrt{5}}{2}\right)^n\right)/\sqrt{5}$; for the case of $k=(n-1)/2$, the number of trace terms is $\left(\left(\frac{1+\sqrt{5}}{2}\right)^{n-1}-\right.$ $\left.\left(\frac{1-\sqrt{5}}{2}\right)^{n-1}\right)/\sqrt{5}$; for the case of $(n, k)=(5t+4, 4t+3)$ or $(5t+1, 4t+1)$ with $t\geq 1$, the number of trace terms is 8; and for the case of $(n, k)=(7t+6, 6t+5)$ or $(7t+1, 6t+1)$ with $t\geq 1$, the number of trace terms is 21. As a byproduct, we find new classes of ternary bent functions with only 8 or 21 trace terms.
- Jul 18 2017 cs.CV arXiv:1707.04993v2Visual signals in a video can be divided into content and motion. While content specifies which objects are in the video, motion describes their dynamics. Based on this prior, we propose the Motion and Content decomposed Generative Adversarial Network (MoCoGAN) framework for video generation. The proposed framework generates a video by mapping a sequence of random vectors to a sequence of video frames. Each random vector consists of a content part and a motion part. While the content part is kept fixed, the motion part is realized as a stochastic process. To learn motion and content decomposition in an unsupervised manner, we introduce a novel adversarial learning scheme utilizing both image and video discriminators. Extensive experimental results on several challenging datasets with qualitative and quantitative comparison to the state-of-the-art approaches, verify effectiveness of the proposed framework. In addition, we show that MoCoGAN allows one to generate videos with same content but different motion as well as videos with different content and same motion.
- Jul 13 2017 cs.CV arXiv:1707.03574v2To provide the possibility of developing objective image quality assessment (IQA) algorithms for THz security images, we constructed the THz security image database (THSID) including a total of 181 THz security images with the resolution of 127*380. The main distortion types in THz security images were first analyzed for the design of subjective evaluation criteria to acquire the mean opinion scores. Subsequently, the existing no-reference IQA algorithms, which were 5 opinion-aware approaches viz., NFERM, GMLF, DIIVINE, BRISQUE and BLIINDS2, and 8 opinion-unaware approaches viz., QAC, SISBLIM, NIQE, FISBLIM, CPBD, S3 and Fish_bb, were executed for the evaluation of the THz security image quality. The statistical results demonstrated the superiority of Fish_bb over the other testing IQA approaches for assessing the THz image quality with PLCC (SROCC) values of 0.8925 (-0.8706), and with RMSE value of 0.3993. The linear regression analysis and Bland-Altman plot further verified that the Fish__bb could substitute for the subjective IQA. Nonetheless, for the classification of THz security images, we tended to use S3 as a criterion for ranking THz security image grades because of the relatively low false positive rate in classifying bad THz image quality into acceptable category (24.69%). Interestingly, due to the specific property of THz image, the average pixel intensity gave the best performance than the above complicated IQA algorithms, with the PLCC, SROCC and RMSE of 0.9001, -0.8800 and 0.3857, respectively. This study will help the users such as researchers or security staffs to obtain the THz security images of good quality. Currently, our research group is attempting to make this research more comprehensive.
- Jul 13 2017 cs.CV arXiv:1707.03816v1Visual tracking is challenging as target objects often undergo significant appearance changes caused by deformation, abrupt motion, background clutter and occlusion. In this paper, we propose to exploit the rich hierarchical features of deep convolutional neural networks to improve the accuracy and robustness of visual tracking. Deep neural networks trained on object recognition datasets consist of multiple convolutional layers. These layers encode target appearance with different levels of abstraction. For example, the outputs of the last convolutional layers encode the semantic information of targets and such representations are invariant to significant appearance variations. However, their spatial resolutions are too coarse to precisely localize the target. In contrast, features from earlier convolutional layers provide more precise localization but are less invariant to appearance changes. We interpret the hierarchical features of convolutional layers as a nonlinear counterpart of an image pyramid representation and explicitly exploit these multiple levels of abstraction to represent target objects. Specifically, we learn adaptive correlation filters on the outputs from each convolutional layer to encode the target appearance. We infer the maximum response of each layer to locate targets in a coarse-to-fine manner. To further handle the issues with scale estimation and target re-detection from tracking failures caused by heavy occlusion or moving out of the view, we conservatively learn another correlation filter that maintains a long-term memory of target appearance as a discriminative classifier. Extensive experimental results on large-scale benchmark datasets show that the proposed algorithm performs favorably against the state-of-the-art tracking methods.
- Jul 13 2017 cs.CR arXiv:1707.03611v1This paper is devoted to measuring the security of cyber networks under advanced persistent threats (APTs). First, an APT-based cyber attack-defense process is modeled as an individual-level dynamical system. Second, the dynamic model is shown to exhibit the global stability. On this basis, a new security metric of cyber networks, which is known as the limit security, is defined as the limit expected fraction of compromised nodes in the networks. Next, the influence of different factors on the limit security is illuminated through theoretical analysis and computer simulation. This work helps understand the security of cyber networks under APTs.
- Jul 11 2017 cs.CR arXiv:1707.02437v2As a new type of cyber attacks, advanced persistent threats (APTs) pose a severe threat to modern society. This paper focuses on the assessment of the risk of APTs. Based on a dynamic model characterizing the time evolution of the state of an organization, the organization's risk is defined as its maximum possible expected loss, and the risk assessment problem is modeled as a constrained optimization problem. The influence of different factors on an organization's risk is uncovered through theoretical analysis. Based on extensive experiments, we speculate that the attack strategy obtained by applying the hill-climbing method to the proposed optimization problem, which we call the HC strategy, always leads to the maximum possible expected loss. We then present a set of five heuristic attack strategies and, through comparative experiments, show that the HC strategy causes a higher risk than all these heuristic strategies do, which supports our conjecture. Finally, the impact of two factors on the attacker's HC cost profit is determined through computer simulations. These findings help understand the risk of APTs in a quantitative manner.
- Jul 11 2017 cs.CV arXiv:1707.02309v1Object tracking is challenging as target objects often undergo drastic appearance changes over time. Recently, adaptive correlation filters have been successfully applied to object tracking. However, tracking algorithms relying on highly adaptive correlation filters are prone to drift due to noisy updates. Moreover, as these algorithms do not maintain long-term memory of target appearance, they cannot recover from tracking failures caused by heavy occlusion or target disappearance in the camera view. In this paper, we propose to learn multiple adaptive correlation filters with both long-term and short-term memory of target appearance for robust object tracking. First, we learn a kernelized correlation filter with an aggressive learning rate for locating target objects precisely. We take into account the appropriate size of surrounding context and the feature representations. Second, we learn a correlation filter over a feature pyramid centered at the estimated target position for predicting scale changes. Third, we learn a complementary correlation filter with a conservative learning rate to maintain long-term memory of target appearance. We use the output responses of this long-term filter to determine if tracking failure occurs. In the case of tracking failures, we apply an incrementally learned detector to recover the target position in a sliding window fashion. Extensive experimental results on large-scale benchmark datasets demonstrate that the proposed algorithm performs favorably against the state-of-the-art methods in terms of efficiency, accuracy, and robustness.
- We study a class of linear network coding (LNC) schemes, called \emphcircular-shift LNC, whose encoding operations at intermediate nodes consist of only circular-shifts and bit-wise addition (XOR). Departing from existing literature, we systematically formulate circular-shift LNC as a special type of vector LNC, where the local encoding kernels of an $L$-dimensional circular-shift linear code of degree $\delta$ are summation of at most $\delta$ cyclic-permutation matrices of size $L$. Under this framework, an intrinsic connection between scalar LNC and circular-shift LNC is established. In particular, on a general network, for some block lengths $L$, every scalar linear solution over GF($2^{L-1}$) can induce an $(L-1, L)$-fractional circular-shift linear solution of degree $(L-1)/2$. Specific to multicast networks, an $(L-1, L)$-fractional circular-shift linear solution of arbitrary degree $\delta$ can be efficiently constructed. With different $\delta$, the constructed solution has an interesting encoding-decoding complexity tradeoff, and when $\delta = (L-1)/2$, it requires fewer binary operations for both encoding and decoding processes compared with scalar LNC. While the constructed solution has one-bit redundancy per edge transmission, we show that this is inevitable, and that circular-shift LNC is insufficient to achieve the exact capacity of some multicast networks. Last, both theoretical and numerical analysis imply that with increasing $L$, a randomly constructed circular-shift linear code has comparable linear solvability behavior to a randomly constructed permutation-based linear code, but has much shorter overheads for random coding.
- Jul 05 2017 cs.CV arXiv:1707.01058v2This work make the first attempt to generate articulated human motion sequence from a single image. On the one hand, we utilize paired inputs including human skeleton information as motion embedding and a single human image as appearance reference, to generate novel motion frames, based on the conditional GAN infrastructure. On the other hand, a triplet loss is employed to pursue appearance-smoothness between consecutive frames. As the proposed framework is capable of jointly exploiting the image appearance space and articulated/kinematic motion space, it generates realistic articulated motion sequence, in contrast to most previous video generation methods which yield blurred motion effects. We test our model on two human action datasets including KTH and Human3.6M, and the proposed framework generates very promising results on both datasets.
- Most existing matching algorithms are one-off algorithms, i.e., they usually measure the distance between the two image feature representation vectors for only one time. In contrast, human's vision system achieves this task, i.e., image matching, by recursively looking at specific/related parts of both images and then making the final judgement. Towards this end, we propose a novel loopy recurrent neural network (Loopy RNN), which is capable of aggregating relationship information of two input images in a progressive/iterative manner and outputting the consolidated matching score in the final iteration. A Loopy RNN features two uniqueness. First, built on conventional long short-term memory (LSTM) nodes, it links the output gate of the tail node to the input gate of the head node, thus it brings up symmetry property required for matching. Second, a monotonous loss designed for the proposed network guarantees increasing confidence during the recursive matching process. Extensive experiments on several image matching benchmarks demonstrate the great potential of the proposed method.
- We present an end-to-end, multimodal, fully convolutional network for extracting semantic structures from document images. We consider document semantic structure extraction as a pixel-wise segmentation task, and propose a unified model that classifies pixels based not only on their visual appearance, as in the traditional page segmentation task, but also on the content of underlying text. Moreover, we propose an efficient synthetic document generation process that we use to generate pretraining data for our network. Once the network is trained on a large set of synthetic documents, we fine-tune the network on unlabeled real documents using a semi-supervised approach. We systematically study the optimum network architecture and show that both our multimodal approach and the synthetic data pretraining significantly boost the performance.
- Jun 08 2017 cs.SI physics.soc-ph arXiv:1706.02035v1This paper addressed the issue of estimating the damage caused by a computer virus. First, an individual-level delayed SIR model capturing the spreading process of a digital virus is derived. Second, the damage inflicted by the virus is modeled as the sum of the economic losses and the cost for developing the antivirus. Next, the impact of different factors, including the delay and the network structure, on the damage is explored by means of computer simulations. Thereby some measures of reducing the damage of a virus are recommended. To our knowledge, this is the first time the antivirus-developing cost is taken into account when estimating the damage of a virus.
- Jun 02 2017 cs.CV arXiv:1706.00212v2Key to automatically generate natural scene images is to properly arrange among various spatial elements, especially in the depth direction. To this end, we introduce a novel depth structure preserving scene image generation network (DSP-GAN), which favors a hierarchical and heterogeneous architecture, for the purpose of depth structure preserving scene generation. The main trunk of the proposed infrastructure is built on a Hawkes point process that models the spatial dependency between different depth layers. Within each layer generative adversarial sub-networks are trained collaboratively to generate realistic scene components, conditioned on the layer information produced by the point process. We experiment our model on a sub-set of SUNdataset with annotated scene images and demonstrate that our models are capable of generating depth-realistic natural scene image.
- May 31 2017 cs.SI arXiv:1705.10618v1Spreading truths and blocking rumors are two typical strategies for inhibiting rumors. In practice, a tradeoff between the two strategies, which is known as the TSRB strategy, may achieve a better cost-effectiveness. This paper is devoted to assessing the effectiveness of the TSRB strategy. For that purpose, an individual-level spreading model (the generic URQT model) capturing the interaction between a rumor and the truth is established. Under the model, a set of criteria for the dying out of a rumor is presented. These criteria capture the combined influence of the basic parameters and the network structures on the effectiveness of the TSRB strategy. Experimental results show that, when the rumor dies out, the dynamics of a simplified URQT model (the linear URQT model) fits well with the actual rumor-truth interacting process. Therefore, the generic URQT model and sometimes the linear URQT model provide a proper basis for assessing the effectiveness of the TSRB strategy.
- May 29 2017 cs.CV arXiv:1705.09467v1Predicting human interaction is challenging as the on-going activity has to be inferred based on a partially observed video. Essentially, a good algorithm should effectively model the mutual influence between the two interacting subjects. Also, only a small region in the scene is discriminative for identifying the on-going interaction. In this work, we propose a relative attention model to explicitly address these difficulties. Built on a tri-coupled deep recurrent structure representing both interacting subjects and global interaction status, the proposed network collects spatio-temporal information from each subject, rectified with global interaction information, yielding effective interaction representation. Moreover, the proposed network also unifies an attention module to assign higher importance to the regions which are relevant to the on-going action. Extensive experiments have been conducted on two public datasets, and the results demonstrate that the proposed relative attention network successfully predicts informative regions between interacting subjects, which in turn yields superior human interaction prediction accuracy.
- Event sequence, asynchronously generated with random timestamp, is ubiquitous among applications. The precise and arbitrary timestamp can carry important clues about the underlying dynamics, and has lent the event data fundamentally different from the time-series whereby series is indexed with fixed and equal time interval. One expressive mathematical tool for modeling event is point process. The intensity functions of many point processes involve two components: the background and the effect by the history. Due to its inherent spontaneousness, the background can be treated as a time series while the other need to handle the history events. In this paper, we model the background by a Recurrent Neural Network (RNN) with its units aligned with time series indexes while the history effect is modeled by another RNN whose units are aligned with asynchronous events to capture the long-range dynamics. The whole model with event type and timestamp prediction output layers can be trained end-to-end. Our approach takes an RNN perspective to point process, and models its background and history effect. For utility, our method allows a black-box treatment for modeling the intensity which is often a pre-defined parametric form in point processes. Meanwhile end-to-end training opens the venue for reusing existing rich techniques in deep network for point process modeling. We apply our model to the predictive maintenance problem using a log dataset by more than 1000 ATMs from a global bank headquartered in North America.
- May 19 2017 cs.SI arXiv:1705.06604v1Spreading truths is recognized as a feasible strategy for inhibiting rumors. This paper is devoted to assessing the effectiveness of the truth-spreading strategy. An individual-level rumor-truth spreading model (the generic URTU model) is derived. Under the model, two criteria for the termination of a rumor are presented. These criteria capture the influence of the network structures on the effectiveness of the truth-spreading strategy. Extensive simulations show that, when the rumor or the truth terminates, the dynamics of a simplified URTU model (the linear URTU model) fits well with the actual rumor-truth interplay process. Therefore, the generic URTU model forms a theoretical basis for assessing the effectiveness of the truth-spreading strategy for restraining rumors.
- May 16 2017 cs.SI arXiv:1705.04818v1The decentralized patch distribution mechanism holds significant promise as an alternative to its centralized counterpart. For the purpose of accurately evaluating the performance of the decentralized patch distribution mechanism and based on the exact SIPS model that accurately captures the average dynamics of the interaction between viruses and patches, a new virus-patch interacting model, which is known as the generic SIPS model, is proposed. This model subsumes the linear SIPS model. The dynamics of the generic SIPS model is studied comprehensively. In particular, a set of criteria for the final extinction or/and long-term survival of viruses or/and patches are presented. Some conditions for the linear SIPS model to accurately capture the average dynamics of the virus-patch interaction are empirically found. As a consequence, the linear SIPS model can be adopted as a standard model for assessing the performance of the distributed patch distribution mechanism, provided the proper conditions are satisfied.
- Apr 25 2017 cs.SI arXiv:1704.06920v1As compared to the traditional advertising, the word-of-mouth (WOM) communications have striking advantages such as significantly lower cost and rapid delivery; this is especially the case with the popularity of online social networks. This paper addresses the issue of maximizing the overall profit of a WOM marketing campaign. A marketing process with both positive and negative WOM is modeled as a dynamical model knwn as the SIPNS model, and the profit maximization problem is modeled as a constrained optimization problem. The influence of different factors on the dynamics of the SIPNS model is revealed experimentally. Also, the impact of different factors on the expected overall profit of a WOM marketing campaign is uncovered experimentally. On this basis, some promotion strategies are suggested. To our knowledge, this is the first time a WOM marketing campaign is treated this way.
- Apr 25 2017 cs.SI arXiv:1704.06910v1This paper addresses the discount pricing in word-of-mouth (WOM) marketing. A new discount strategy known as the Infection-Based Discount (IBD) strategy is proposed. The basic idea of the IBD strategy lies in that each customer enjoys a discount that is linearly proportional to his/her influence in the WOM network. To evaluate the performance of the IBD strategy, the WOM spreading process is modeled as a dynamic model known as the DPA model, and the performance of the IBD strategy is modeled as a function of the basic discount. Next, the influence of different factors, including the basic discount and the WOM network, on the dynamics of the DPA model is revealed experimentally. Finally, the influence of different factors on the performance of the IBD strategy is uncovered experimentally. On this basis, some promotional measures are recommended.
- Apr 21 2017 cs.CV arXiv:1704.06020v2Despite the promising progress made in recent years, person re-identification (re-ID) remains a challenging task due to the complex variations in human appearances from different camera views. For this challenging problem, a large variety of algorithms have been developed in the fully-supervised setting, requiring access to a large amount of labeled training data. However, the main bottleneck for fully-supervised re-ID is the limited availability of labeled training samples. To address this problem, in this paper, we propose a self-trained subspace learning paradigm for person re-ID which effectively utilizes both labeled and unlabeled data to learn a discriminative subspace where person images across disjoint camera views can be easily matched. The proposed approach first constructs pseudo pairwise relationships among unlabeled persons using the k-nearest neighbors algorithm. Then, with the pseudo pairwise relationships, the unlabeled samples can be easily combined with the labeled samples to learn a discriminative projection by solving an eigenvalue problem. In addition, we refine the pseudo pairwise relationships iteratively, which further improves the learning performance. A multi-kernel embedding strategy is also incorporated into the proposed approach to cope with the non-linearity in person's appearance and explore the complementation of multiple kernels. In this way, the performance of person re-ID can be greatly enhanced when training data are insufficient. Experimental results on six widely-used datasets demonstrate the effectiveness of our approach and its performance can be comparable to the reported results of most state-of-the-art fully-supervised methods while using much fewer labeled data.
- Apr 19 2017 cs.NE arXiv:1704.05174v1Optimization techniques play an important role in several scientific and real-world applications, thus becoming of great interest for the community. As a consequence, a number of open-source libraries are available in the literature, which ends up fostering the research and development of new techniques and applications. In this work, we present a new library for the implementation and fast prototyping of nature-inspired techniques called LibOPT. Currently, the library implements 15 techniques and 112 benchmarking functions, as well as it also supports 11 hypercomplex-based optimization approaches, which makes it one of the first of its kind. We showed how one can easily use and also implement new techniques in LibOPT under the C paradigm. Examples are provided with samples of source-code using benchmarking functions.
- We propose a novel fifth-generation (5G) rapid prototyping (RaPro) system architecture by combining FPGA-privileged modules from a software defined radio (or FPGA-coprocessor) and high-level programming language for advanced algorithms from multi-core general purpose processors. The proposed system architecture exhibits excellent flexibility and scalability in the development of a 5G prototyping system. As a proof of concept, a multi-user full-dimension multiple-input and multiple-output system is established based on the proposed architecture. Experimental results demonstrate the superiority of the proposed architecture in large-scale antenna and wideband communication systems.
- Apr 18 2017 cs.RO arXiv:1704.05016v1Loop closure detection (LCD) is an indispensable part of simultaneous localization and mapping systems (SLAM); it enables robots to produce a consistent map by recognizing previously visited places. When robots operate over extended periods, robustness to viewpoint and condition changes as well as satisfactory real-time performance become essential requirements for a practical LCD system. This paper presents an approach to directly utilize the outputs at the intermediate layer of a pre-trained convolutional neural network (CNN) as image descriptors. The matching location is determined by matching the image sequences through a method called SeqCNNSLAM. The utility of SeqCNNSLAM is comprehensively evaluated in terms of viewpoint and condition invariance. Experiments show that SeqCNNSLAM outperforms state-of-the-art LCD systems, such as SeqSLAM and Change Removal, in most cases. To allow for the real-time performance of SeqCNNSLAM, an acceleration method, A-SeqCNNSLAM, is established. This method exploits the location relationship between the matching images of adjacent images to reduce the matching range of the current image. Results demonstrate that acceleration of 4-6 is achieved with minimal accuracy degradation, and the method's runtime satisfies the real-time demand. To extend the applicability of A-SeqCNNSLAM to new environments, a method called O-SeqCNNSLAM is established for the online adjustment of the parameters of A-SeqCNNSLAM.
- Apr 12 2017 cs.CV arXiv:1704.03152v1In recent years, Deep Learning has been successfully applied to multimodal learning problems, with the aim of learning useful joint representations in data fusion applications. When the available modalities consist of time series data such as video, audio and sensor signals, it becomes imperative to consider their temporal structure during the fusion process. In this paper, we propose the Correlational Recurrent Neural Network (CorrRNN), a novel temporal fusion model for fusing multiple input modalities that are inherently temporal in nature. Key features of our proposed model include: (i) simultaneous learning of the joint representation and temporal dependencies between modalities, (ii) use of multiple loss terms in the objective function, including a maximum correlation loss term to enhance learning of cross-modal information, and (iii) the use of an attention model to dynamically adjust the contribution of different input modalities to the joint representation. We validate our model via experimentation on two different tasks: video- and sensor-based activity classification, and audio-visual speech recognition. We empirically analyze the contributions of different components of the proposed CorrRNN model, and demonstrate its robustness, effectiveness and state-of-the-art performance on multiple datasets.
- Apr 04 2017 cs.CV arXiv:1704.00036v1Registration involving one or more images containing pathologies is challenging, as standard image similarity measures and spatial transforms cannot account for common changes due to pathologies. Low-rank/Sparse (LRS) decomposition removes pathologies prior to registration; however, LRS is memory-demanding and slow, which limits its use on larger data sets. Additionally, LRS blurs normal tissue regions, which may degrade registration performance. This paper proposes an efficient alternative to LRS: (1) normal tissue appearance is captured by principal component analysis (PCA) and (2) blurring is avoided by an integrated model for pathology removal and image reconstruction. Results on synthetic and BRATS 2015 data demonstrate its utility.
- Apr 03 2017 cs.CV arXiv:1703.10902v1We introduce a deep encoder-decoder architecture for image deformation prediction from multimodal images. Specifically, we design an image-patch-based deep network that jointly (i) learns an image similarity measure and (ii) the relationship between image patches and deformation parameters. While our method can be applied to general image registration formulations, we focus on the Large Deformation Diffeomorphic Metric Mapping (LDDMM) registration model. By predicting the initial momentum of the shooting formulation of LDDMM, we preserve its mathematical properties and drastically reduce the computation time, compared to optimization-based approaches. Furthermore, we create a Bayesian probabilistic version of the network that allows evaluation of registration uncertainty via sampling of the network at test time. We evaluate our method on a 3D brain MRI dataset using both T1- and T2-weighted images. Our experiments show that our method generates accurate predictions and that learning the similarity measure leads to more consistent registrations than relying on generic multimodal image similarity measures, such as mutual information. Our approach is an order of magnitude faster than optimization-based LDDMM.
- Apr 03 2017 cs.CV arXiv:1703.10908v4This paper introduces Quicksilver, a fast deformable image registration method. Quicksilver registration for image-pairs works by patch-wise prediction of a deformation model based directly on image appearance. A deep encoder-decoder network is used as the prediction model. While the prediction strategy is general, we focus on predictions for the Large Deformation Diffeomorphic Metric Mapping (LDDMM) model. Specifically, we predict the momentum-parameterization of LDDMM, which facilitates a patch-wise prediction strategy while maintaining the theoretical properties of LDDMM, such as guaranteed diffeomorphic mappings for sufficiently strong regularization. We also provide a probabilistic version of our prediction network which can be sampled during the testing time to calculate uncertainties in the predicted deformations. Finally, we introduce a new correction network which greatly increases the prediction accuracy of an already existing prediction network. We show experimental results for uni-modal atlas-to-image as well as uni- / multi- modal image-to-image registrations. These experiments demonstrate that our method accurately predicts registrations obtained by numerical optimization, is very fast, achieves state-of-the-art registration results on four standard validation datasets, and can jointly learn an image similarity measure. Quicksilver is freely available as an open-source software.
- Mar 27 2017 cs.LG arXiv:1703.08524v1A variety of real-world processes (over networks) produce sequences of data whose complex temporal dynamics need to be studied. More especially, the event timestamps can carry important information about the underlying network dynamics, which otherwise are not available from the time-series evenly sampled from continuous signals. Moreover, in most complex processes, event sequences and evenly-sampled times series data can interact with each other, which renders joint modeling of those two sources of data necessary. To tackle the above problems, in this paper, we utilize the rich framework of (temporal) point processes to model event data and timely update its intensity function by the synergic twin Recurrent Neural Networks (RNNs). In the proposed architecture, the intensity function is synergistically modulated by one RNN with asynchronous events as input and another RNN with time series as input. Furthermore, to enhance the interpretability of the model, the attention mechanism for the neural point process is introduced. The whole model with event type and timestamp prediction output layers can be trained end-to-end and allows a black-box treatment for modeling the intensity. We substantiate the superiority of our model in synthetic data and three real-world benchmark datasets.
- Mar 06 2017 cs.CV arXiv:1703.01025v1In this study, a multi-task deep neural network is proposed for skin lesion analysis. The proposed multi-task learning model solves different tasks (e.g., lesion segmentation and two independent binary lesion classifications) at the same time by exploiting commonalities and differences across tasks. This results in improved learning efficiency and potential prediction accuracy for the task-specific models, when compared to training the individual models separately. The proposed multi-task deep learning model is trained and evaluated on the dermoscopic image sets from the International Skin Imaging Collaboration (ISIC) 2017 Challenge - Skin Lesion Analysis towards Melanoma Detection, which consists of 2000 training samples and 150 evaluation samples. The experimental results show that the proposed multi-task deep learning model achieves promising performances on skin lesion segmentation and classification. The average value of Jaccard index for lesion segmentation is 0.724, while the average values of area under the receiver operating characteristic curve (AUC) on two individual lesion classifications are 0.880 and 0.972, respectively.
- Robot awareness of human actions is an essential research problem in robotics with many important real-world applications, including human-robot collaboration and teaming. Over the past few years, depth sensors have become a standard device widely used by intelligent robots for 3D perception, which can also offer human skeletal data in 3D space. Several methods based on skeletal data were designed to enable robot awareness of human actions with satisfactory accuracy. However, previous methods treated all body parts and features equally important, without the capability to identify discriminative body parts and features. In this paper, we propose a novel simultaneous Feature And Body-part Learning (FABL) approach that simultaneously identifies discriminative body parts and features, and efficiently integrates all available information together to enable real-time robot awareness of human behaviors. We formulate FABL as a regression-like optimization problem with structured sparsity-inducing norms to model interrelationships of body parts and features. We also develop an optimization algorithm to solve the formulated problem, which possesses a theoretical guarantee to find the optimal solution. To evaluate FABL, three experiments were performed using public benchmark datasets, including the MSR Action3D and CAD-60 datasets, as well as a Baxter robot in practical assistive living applications. Experimental results show that our FABL approach obtains a high recognition accuracy with a processing speed of the order-of-magnitude of 10e4 Hz, which makes FABL a promising method to enable real-time robot awareness of human behaviors in practical robotics applications.