results for au:Zhang_B in:cs

- Nov 15 2017 cs.CV arXiv:1711.04192v1Correlation filters are special classifiers designed for shift-invariant object recognition, which are robust to pattern distortions. The recent literature shows that combining a set of sub-filters trained based on a single or a small group of images obtains the best performance. The idea is equivalent to estimating variable distribution based on the data sampling (bagging), which can be interpreted as finding solutions (variable distribution approximation) directly from sampled data space. However, this methodology fails to account for the variations existed in the data. In this paper, we introduce an intermediate step -- solution sampling -- after the data sampling step to form a subspace, in which an optimal solution can be estimated. More specifically, we propose a new method, named latent constrained correlation filters (LCCF), by mapping the correlation filters to a given latent subspace, and develop a new learning framework in the latent subspace that embeds distribution-related constraints into the original problem. To solve the optimization problem, we introduce a subspace based alternating direction method of multipliers (SADMM), which is proven to converge at the saddle point. Our approach is successfully applied to three different tasks, including eye localization, car detection and object tracking. Extensive experiments demonstrate that LCCF outperforms the state-of-the-art methods. The source code will be publicly available. https://github.com/bczhangbczhang/.
- In this paper, we derive a temporal arbitrage policy for storage via reinforcement learning. Real-time price arbitrage is an important source of revenue for storage units, but designing good strategies have proven to be difficult because of the highly uncertain nature of the prices. Instead of current model predictive or dynamic programming approaches, we use reinforcement learning to design a two-thresholds policy. This policy is learned through repeated charge and discharge actions performed by the storage unit through updating a value matrix. We design a reward function that does not only reflect the instant profit of charge/discharge decisions but also incorporate the history information. Simulation results demonstrate our designed reward function leads to significant performance improvement compared with existing algorithms.
- Nov 09 2017 cs.DC arXiv:1711.02976v1RPYFMM is a software package for the efficient evaluation of the potential field governed by the Rotne-Prager-Yamakawa (RPY) tensor interactions in biomolecular hydrodynamics simulations. In our algorithm, the RPY tensor is decomposed as a linear combination of four Laplace interactions, each of which is evaluated using the adaptive fast multipole method (FMM) [1] where the exponential expansions are applied to diagonalize the multipole-to-local translation operators. RPYFMM offers a unified execution on both shared and distributed memory computers by leveraging the DASHMM library [2, 3]. Preliminary numerical results show that the interactions for a molecular system of 15 million particles (beads) can be computed within one second on a Cray XC30 cluster using 12, 288 cores, while achieving approximately 54% strong-scaling efficiency.
- Distributed energy resources (DERs) can serve as non-wire alternatives to capacity expansion by managing peak load to avoid or defer traditional expansion projects. In this paper, we study a planning problem that co-optimizes DERs investment and operation (e.g., energy efficiency, energy storage, demand response, solar photovoltaic) and the timing of capacity expansion. We show that this problem can be written as a convex program. However, the problem potentially includes millions of variables because we model the operation of DERs over decades. We use the Dantzig-Wolfe Decomposition to solve this high-dimensional, non-linear problem. Finally, we present a real planning problem at the University of Washington Seattle Campus.
- The paper proposes an inductive semi-supervised learning method, called Smooth Neighbors on Teacher Graphs (SNTG). At each iteration during training, a graph is dynamically constructed based on predictions of the teacher model, i.e., the implicit self-ensemble of models. Then the graph serves as a similarity measure with respect to which the representations of "similar" neighboring points are learned to be smooth on the low dimensional manifold. We achieve state-of-the-art results on semi-supervised learning benchmarks. The error rates are 9.89%, 3.99% for CIFAR-10 with 4000 labels, SVHN with 500 labels, respectively. In particular, the improvements are significant when the labels are scarce. For non-augmented MNIST with only 20 labels, the error rate is reduced from previous 4.81% to 1.36%. Our method is also effective under noisy supervision and shows robustness to incorrect labels.
- Oct 18 2017 cs.DC arXiv:1710.06316v1We present an updated version of the AFMPB package for fast calculation of molecular solvation-free energy. The main feature of the new version is the successful adoption of the DASHMM library, which enables AFMPB to operate on distributed memory computers. As a result, the new version can easily handle larger molecules or situations with higher accuracy requirements. To demonstrate the updated code, we applied the new version to a dengue virus system with more than one million atoms and a mesh with approximately 20 million triangles, and were able to reduce the time-to-solution from 10 hours reported in the previous release on a shared memory computer to less than 30 seconds on a Cray XC30 cluster using 12, 288 cores.
- Sep 26 2017 cs.GT arXiv:1709.08441v3We study the equilibrium quality under user uncertainty in a multi-commodity selfish routing game with many types of users, where each user type experiences a different level of uncertainty. We consider a new model of uncertainty where each user-type over or under-estimates their congestion costs by a multiplicative constant. We present a variety of theoretical results showing that when users under-estimate their costs, the network congestion decreases at equilibrium, whereas over-estimation of costs leads to increased equilibrium congestion. Motivated by applications in urban transportation networks, we perform simulations consisting of parking users and through traffic on synthetic and realistic network topologies. In light of the dynamic pricing policies adopted by network operators to tackle congestion, our results indicate that while users' perception of these prices can significantly impact the policy's efficacy, optimism in the face of uncertainty leads to favorable network conditions.
- We study the optimal control of battery energy storage under a general "pay-for-performance" setup such as providing frequency regulation and renewable integration. Batteries need to carefully balance the trade-off between following to the instruction signals and their degradation costs in real-time. Existing battery control strategies either do not consider the uncertainty of future signals, or cannot accurately account for battery cycle aging mechanism during operation. In this work, we take a different approach to the optimal battery control problem. Instead of attacking the complexity of battery degradation function or the lack of future information one at a time, we address these two challenges together in a joint fashion. In particular, we present an electrochemically accurate and trackable battery degradation model called the rainflow cycle-based model. We prove the degradation cost is convex. Then we propose an online control policy with a simple threshold structure and show it achieve near-optimal performance with respect to an offline controller that has complete future information. We explicitly characterize the optimality gap and show it is independent to length of the time of operations. Simulation results with both synthetic and real regulation traces are conducted to illustrate the theoretical results.
- Sep 13 2017 cs.SY arXiv:1709.03746v1To settle a large-scale integration of renewable distributed generations (DGs), it requires to assess the maximal DG hosting capacity of active distribution networks (ADNs). For fully exploiting the ability of ADNs to accommodate DG, this paper proposes a robust comprehensive DG capacity assessment method considering three-phase power flow modelling and active network management (ANM) techniques. The two-stage adjustable robust optimization is employed to tackle the uncertainties of load demands and DG outputs. With our method, system planners can obtain the maximum penetration level of DGs with their optimal sizing and sitting decisions. Meanwhile, the robust optimal ANM schemes can be generated for each operation time period, including network reconfiguration, on-load-tap-changers regulation, and reactive power compensation. In addition, a three-step optimization algorithm is proposed to enhance the accuracy of DG capacity assessment results. The optimality and robustness of our method are validated via numerical tests on an unbalanced IEEE 33-bus distribution system.
- Power systems are increasingly operated in corrective rather than preventive security mode, which means that appropriate control actions must be taken immediately after a contingency has occurred. This paper proposes an on-line algorithm for automatically alleviating contingencies such as voltage limit violations and line overloads. Unlike previously proposed approaches, the network itself serves as a natural solver of the power flow equations. This makes it possible to start the implementation immediately and avoids problems caused by modeling errors. Every time the controller receives measurements from the grid, it evaluates the presence of contingencies, and computes the optimal corrective actions that can be implemented before the next sampling period, subject to ramping constraints of the generators. These corrective actions are implemented through the standard Automatic Generation Control. Finding the optimal incremental corrective actions is fast because this problem is linearized. The effectiveness of this algorithm at correcting both line overloads and voltage violations is demonstrated using the IEEE-118 Bus test system.
- Aug 28 2017 cs.CV arXiv:1708.07590v2Recurrent Neural Networks (RNNs) have been widely used in natural language processing and computer vision. Among them, the Hierarchical Multi-scale RNN (HM-RNN), a kind of multi-scale hierarchical RNN proposed recently, can learn the hierarchical temporal structure from data automatically. In this paper, we extend the work to solve the computer vision task of action recognition. However, in sequence-to-sequence models like RNN, it is normally very hard to discover the relationships between inputs and outputs given static inputs. As a solution, attention mechanism could be applied to extract the relevant information from input thus facilitating the modeling of input-output relationships. Based on these considerations, we propose a novel attention network, namely Hierarchical Multi-scale Attention Network (HM-AN), by combining the HM-RNN and the attention mechanism and apply it to action recognition. A newly proposed gradient estimation method for stochastic neurons, namely Gumbel-softmax, is exploited to implement the temporal boundary detectors and the stochastic hard attention mechanism. To amealiate the negative effect of sensitive temperature of the Gumbel-softmax, an adaptive temperature training method is applied to better the system performance. The experimental results demonstrate the improved effect of HM-AN over LSTM with attention on the vision task. Through visualization of what have been learnt by the networks, it can be observed that both the attention regions of images and the hierarchical temporal structure can be captured by HM-AN.
- We consider a wide range of regularized stochastic minimization problems with two regularization terms, one of which is composed with a linear function. This optimization model abstracts a number of important applications in artificial intelligence and machine learning, such as fused Lasso, fused logistic regression, and a class of graph-guided regularized minimization. The computational challenges of this model are in two folds. On one hand, the closed-form solution of the proximal mapping associated with the composed regularization term or the expected objective function is not available. On the other hand, the calculation of the full gradient of the expectation in the objective is very expensive when the number of input data samples is considerably large. To address these issues, we propose a stochastic variant of extra-gradient type methods, namely \textsfStochastic Primal-Dual Proximal ExtraGradient descent (SPDPEG), and analyze its convergence property for both convex and strongly convex objectives. For general convex objectives, the uniformly average iterates generated by \textsfSPDPEG converge in expectation with $O(1/\sqrt{t})$ rate. While for strongly convex objectives, the uniformly and non-uniformly average iterates generated by \textsfSPDPEG converge with $O(\log(t)/t)$ and $O(1/t)$ rates, respectively. The order of the rate of the proposed algorithm is known to match the best convergence rate for first-order stochastic algorithms. Experiments on fused logistic regression and graph-guided regularized logistic regression problems show that the proposed algorithm performs very efficiently and consistently outperforms other competing algorithms.
- Aug 16 2017 cs.IR arXiv:1708.04531v1The name disambiguation task partitions a collection of records pertaining to a given name, such that there is a one-to-one correspondence between the partitions and a group of people, all sharing that given name. Most existing solutions for this task are proposed for static data. However, more realistic scenarios stipulate emergence of records in a streaming fashion where records may belong to known as well as unknown persons all sharing the same name. This requires a flexible name disambiguation algorithm that can not only classify records of known persons represented in the train- ing data by their existing records but can also identify records of new ambiguous persons with no existing records included in the initial training dataset. Toward achieving this objective, in this paper we propose a Bayesian non-exhaustive classification frame- work for solving online name disambiguation. In particular, we present a Dirichlet Process Gaussian Mixture Model (DPGMM) as a core engine for online name disambiguation task. Meanwhile, two online inference algorithms, namely one-pass Gibbs sampler and Sequential Importance Sampling with Resampling (also known as particle filtering), are proposed to simultaneously perform online classification and new class discovery. As a case study we consider bibliographic data in a temporal stream format and disambiguate authors by partitioning their papers into homogeneous groups.Our experimental results demonstrate that the proposed method is significantly better than existing methods for performing online name disambiguation task. We also propose an interactive version of our online name disambiguation method designed to leverage user feedback to improve prediction accuracy.
- Aug 08 2017 cs.CV arXiv:1708.01846v1Low-rank decomposition (LRD) is a state-of-the-art method for visual data reconstruction and modelling. However, it is a very challenging problem when the image data contains significant occlusion, noise, illumination variation, and misalignment from rotation or viewpoint changes. We leverage the specific structure of data in order to improve the performance of LRD when the data are not ideal. To this end, we propose a new framework that embeds manifold priors into LRD. To implement the framework, we design an alternating direction method of multipliers (ADMM) method which efficiently integrates the manifold constraints during the optimization process. The proposed approach is successfully used to calculate low-rank models from face images, hand-written digits and planar surface images. The results show a consistent increase of performance when compared to the state-of-the-art over a wide range of realistic image misalignments and corruptions.
- An important function of aggregators is to enable the participation of small energy storage units in electricity markets. This paper studies two generally overlooked aspects related to aggregators of energy storage: i) the relationship between the aggregator and its constituent storage units and ii) the aggregator's effect on system welfare. Regarding i), we show that short-term outcomes can be Pareto-inefficient: all players could be better-off. In practice, however, aggregators and storage units are likely to engage in long rather than short-term relationships. Using Nash Bargaining Theory, we show that aggregators and storage units are likely to cooperate in the long-term. A rigorous understanding of the aggregator-storage unit relationship is fundamental to model the aggregator's participation in the market. Regarding ii), we first show that a profit-seeking energy storage aggregator is always beneficial to the system when compared to a system without storage, regardless of size or market power the aggregator may have. However, due to market power, a monopolist aggregator may act in a socially suboptimal manner. We propose a pricing scheme designed to mitigate market power abuse by the aggregator. This pricing scheme has several important characteristics: its formulation requires no private information, it incentivizes a rational aggregator to behave in a socially optimal manner, and allows for regulation of the aggregator's profit.
- Traditional Recurrent Neural Networks assume vectorized data as inputs. However many data from modern science and technology come in certain structures such as tensorial time series data. To apply the recurrent neural networks for this type of data, a vectorisation process is necessary, while such a vectorisation leads to the loss of the precise information of the spatial or longitudinal dimensions. In addition, such a vectorized data is not an optimum solution for learning the representation of the longitudinal data. In this paper, we propose a new variant of tensorial neural networks which directly take tensorial time series data as inputs. We call this new variant as Tensorial Recurrent Neural Network (TRNN). The proposed TRNN is based on tensor Tucker decomposition.
- Aug 01 2017 cs.SD arXiv:1707.09890v1Traditional intelligent fault diagnosis of rolling bearings work well only under a common assumption that the labeled training data (source domain) and unlabeled testing data (target domain) are drawn from the same distribution. When the distribution changes, most fault diagnosis models need to be rebuilt from scratch using newly recollected labeled training data. However, it is expensive or impossible to annotate huge amount of training data to rebuild such new model. Meanwhile, large amounts of labeled training data have not been fully utilized yet, which is apparently a waste of resources. As one of the important research directions of transfer learning, domain adaptation (DA) typically aims at minimizing the differences between distributions of different domains in order to minimize the cross-domain prediction error by taking full advantage of information coming from both source and target domains. In this paper, we present one of the first studies on unsupervised DA in the field of fault diagnosis of rolling bearings under varying working conditions and a novel diagnosis strategy based on unsupervised DA using subspace alignment (SA) is proposed. After processed by unsupervised DA with SA, the distributions of training data and testing data become close and the classifier trained on training data can be used to classify the testing data. Experimental results on the 60 domain adaptation diagnosis problems under varying working condition in Case Western Reserve benchmark data and 12 domain adaptation diagnosis problems under varying working conditions in our new data are given to demonstrate the effectiveness of the proposed method. The proposed methods can effectively distinguish not only bearing faults categories but also fault severities.
- Scenario generation is an important step in the operation and planning of power systems with high renewable penetrations. In this work, we proposed a data-driven approach for scenario generation using generative adversarial networks, which is based on two interconnected deep neural networks. Compared with existing methods based on probabilistic models that are often hard to scale or sample from, our method is data-driven, and captures renewable energy production patterns in both temporal and spatial dimensions for a large number of correlated resources. For validation, we use wind and solar times-series data from NREL integration data sets. We demonstrate that the proposed method is able to generate realistic wind and photovoltaic power profiles with full diversity of behaviors. We also illustrate how to generate scenarios based on different conditions of interest by using labeled data during training. For example, scenarios can be conditioned on weather events~(e.g. high wind day) or time of the year~(e,g. solar generation for a day in July). Because of the feedforward nature of the neural networks, scenarios can be generated extremely efficiently without sophisticated sampling techniques.
- Jul 27 2017 cs.CV arXiv:1707.08289v1Image matting plays an important role in image and video editing. However, the formulation of image matting is inherently ill-posed. Traditional methods usually employ interaction to deal with the image matting problem with trimaps and strokes, and cannot run on the mobile phone in real-time. In this paper, we propose a real-time automatic deep matting approach for mobile devices. By leveraging the densely connected blocks and the dilated convolution, a light full convolutional network is designed to predict a coarse binary mask for portrait images. And a feathering block, which is edge-preserving and matting adaptive, is further developed to learn the guided filter and transform the binary mask into alpha matte. Finally, an automatic portrait animation system based on fast deep matting is built on mobile devices, which does not need any interaction and can realize real-time matting with 15 fps. The experiments show that the proposed approach achieves comparable results with the state-of-the-art matting solvers.
- Jul 25 2017 cs.CV arXiv:1707.07411v1Traffic scene recognition is an important and challenging issue in Intelligent Transportation Systems (ITS). Recently, Convolutional Neural Network (CNN) models have achieved great success in many applications, including scene classification. The remarkable representational learning capability of CNN remains to be further explored for solving real-world problems. Vector of Locally Aggregated Descriptors (VLAD) encoding has also proved to be a powerful method in catching global contextual information. In this paper, we attempted to solve the traffic scene recognition problem by combining the features representational capabilities of CNN with the VLAD encoding scheme. More specifically, the CNN features of image patches generated by a region proposal algorithm are encoded by applying VLAD, which subsequently represent an image in a compact representation. To catch the spatial information, spatial pyramids are exploited to encode CNN features. We experimented with a dataset of 10 categories of traffic scenes, with satisfactory categorization performances.
- Jul 13 2017 cs.CV arXiv:1707.03692v1Gesture recognition is a challenging problem in the field of biometrics. In this paper, we integrate Fisher criterion into Bidirectional Long-Short Term Memory (BLSTM) network and Bidirectional Gated Recurrent Unit (BGRU),thus leading to two new deep models termed as F-BLSTM and F-BGRU. BothFisher discriminative deep models can effectively classify the gesture based on analyzing the acceleration and angular velocity data of the human gestures. Moreover, we collect a large Mobile Gesture Database (MGD) based on the accelerations and angular velocities containing 5547 sequences of 12 gestures. Extensive experiments are conducted to validate the superior performance of the proposed networks as compared to the state-of-the-art BLSTM and BGRU on MGD database and two benchmark databases (i.e. BUAA mobile gesture and SmartWatch gesture).
- Jul 11 2017 cs.CV arXiv:1707.02406v1In this paper, a level-wise mixture model (LMM) is developed by embedding visual hierarchy with deep networks to support large-scale visual recognition (i.e., recognizing thousands or even tens of thousands of object classes), and a Bayesian approach is used to adapt a pre-trained visual hierarchy automatically to the improvements of deep features (that are used for image and object class representation) when more representative deep networks are learned along the time. Our LMM model can provide an end-to-end approach for jointly learning: (a) the deep networks to extract more discriminative deep features for image and object class representation; (b) the tree classifier for recognizing large numbers of object classes hierarchically; and (c) the visual hierarchy adaptation for achieving more accurate indexing of large numbers of object classes hierarchically. By supporting joint learning of the tree classifier, the deep networks and the visual hierarchy adaptation, our LMM algorithm can provide an effective approach for controlling inter-level error propagation effectively, thus it can achieve better accuracy rates on large-scale visual recognition. Our experiments are carried on ImageNet1K and ImageNet10K image sets, and our LMM algorithm can achieve very competitive results on both the accuracy rates and the computation efficiency as compared with the baseline methods.
- Question-answering (QA) on video contents is a significant challenge for achieving human-level intelligence as it involves both vision and language in real-world settings. Here we demonstrate the possibility of an AI agent performing video story QA by learning from a large amount of cartoon videos. We develop a video-story learning model, i.e. Deep Embedded Memory Networks (DEMN), to reconstruct stories from a joint scene-dialogue video stream using a latent embedding space of observed data. The video stories are stored in a long-term memory component. For a given question, an LSTM-based attention model uses the long-term memory to recall the best question-story-answer triplet by focusing on specific words containing key information. We trained the DEMN on a novel QA dataset of children's cartoon video series, Pororo. The dataset contains 16,066 scene-dialogue pairs of 20.5-hour videos, 27,328 fine-grained sentences for scene description, and 8,913 story-related QA pairs. Our experimental results show that the DEMN outperforms other QA models. This is mainly due to 1) the reconstruction of video stories in a scene-dialogue combined form that utilize the latent embedding and 2) attention. DEMN also achieved state-of-the-art results on the MovieQA benchmark.
- Jul 05 2017 cs.LG arXiv:1707.00802v2We describe a parallel bayesian online deep learning framework (PBODL) for click-through rate (CTR) prediction within today's Tencent advertising system, which provides quick and accurate learning of user preferences. We first explain the framework with a deep probit regression model, which is trained with probabilistic back-propagation in the mode of assumed Gaussian density filtering. Then we extend the model family to a variety of bayesian online models with increasing feature embedding capabilities, such as Sparse-MLP, FM-MLP and FFM-MLP. Finally, we implement a parallel training system based on a stream computing infrastructure and parameter servers. Experiments with public available datasets and Tencent industrial datasets show that models within our framework perform better than several common online models, such as AdPredictor, FTRL-Proximal and MatchBox. Online A/B test within Tencent advertising system further proves that our framework could achieve CTR and CPM lift by learning more quickly and accurately.
- This paper presents a Semantic Attribute Modulation (SAM) for language modeling and style variation. The semantic attribute modulation includes various document attributes, such as titles, authors, and document categories. We consider two types of attributes, (title attributes and category attributes), and a flexible attribute selection scheme by automatically scoring them via an attribute attention mechanism. The semantic attributes are embedded into the hidden semantic space as the generation inputs. With the attributes properly harnessed, our proposed SAM can generate interpretable texts with regard to the input attributes. Qualitative analysis, including word semantic analysis and attention values, shows the interpretability of SAM. On several typical text datasets, we empirically demonstrate the superiority of the Semantic Attribute Modulated language model with different combinations of document attributes. Moreover, we present a style variation for the lyric generation using SAM, which shows a strong connection between the style variation and the semantic attributes.
- Demand response aims to stimulate electricity consumers to modify their loads at critical time periods. In this paper, we consider signals in demand response programs as a binary treatment to the customers and estimate the average treatment effect, which is the average change in consumption under the demand response signals. More specifically, we propose to estimate this effect by linear regression models and derive several estimators based on the different models. From both synthetic and real data, we show that including more information about the customers does not always improve estimation accuracy: the interaction between the side information and the demand response signal must be carefully modeled. In addition, we compare the traditional linear regression model with the modified covariate method which models the interaction between treatment effect and covariates. We analyze the variances of these estimators and discuss different cases where each respective estimator works the best. The purpose of these comparisons is not to claim the superiority of the different methods, rather we aim to provide practical guidance on the most suitable estimator to use under different settings. Our results are validated using data collected by Pecan Street and EnergyPlus.
- A grand challenge for power grid management lies in how to plan and operate with increasing penetration of distributed energy resources (DERs), such as solar photovoltaics and electric vehicles, which disturb the power grid stability. Existing approaches are unable to verify if a point is on a loadability boundary or characterize all loadability boundary points exactly. This inability leads to a poor understanding of locational hosting capacity for accommodating distributed resources. To solve these problems, we compare existing approaches and propose a rectangular coordinate-based analysis, which drew less attention in the past. We demonstrate that such a coordinate (1) provides an integrated geometric understanding of active and reactive power flow equations, (2) enables linear representation of elements in the Jacobian matrix, (3) verifies if an operating point is on the loadability boundary and what is the margin, and ($4$) characterizes the power flow feasibility boundary points. Finally, IEEE standard test cases demonstrate the capability of the new method.
- As a cutting-edge technology, microgrids feature intelligent EMSs and sophisticated control, which will dramatically change our energy infrastructure. The modern microgrids are a relatively recent development with high potential to bring distributed generation, DES devices, controllable loads, communication infrastructure, and many new technologies into the mainstream. As a more controllable and intelligent entity, a microgrid has more growth potential than ever before. However, there are still many open questions, such as the future business models and economics. What is the cost-benefit to the end-user? How should we systematically evaluate the potential benefits and costs of control and energy management in a microgrid?
- Advances in artificial intelligence (AI) will transform modern life by reshaping transportation, health, science, finance, and the military. To adapt public policy, we need to better anticipate these advances. Here we report the results from a large survey of machine learning researchers on their beliefs about progress in AI. Researchers predict AI will outperform humans in many activities in the next ten years, such as translating languages (by 2024), writing high-school essays (by 2026), driving a truck (by 2027), working in retail (by 2031), writing a bestselling book (by 2049), and working as a surgeon (by 2053). Researchers believe there is a 50% chance of AI outperforming humans in all tasks in 45 years and of automating all human jobs in 120 years, with Asian respondents expecting these dates much sooner than North Americans. These results will inform discussion amongst researchers and policymakers about anticipating and managing trends in AI.
- May 22 2017 cs.CV arXiv:1705.06950v1We describe the DeepMind Kinetics human action video dataset. The dataset contains 400 human action classes, with at least 400 video clips for each action. Each clip lasts around 10s and is taken from a different YouTube video. The actions are human focussed and cover a broad range of classes including human-object interactions such as playing instruments, as well as human-human interactions such as shaking hands. We describe the statistics of the dataset, how it was collected, and give some baseline performance figures for neural network architectures trained and tested for human action classification on this dataset. We also carry out a preliminary analysis of whether imbalance in the dataset leads to bias in the classifiers.
- May 10 2017 cs.CV arXiv:1705.03146v2Recently, the soft attention mechanism, which was originally proposed in language processing, has been applied in computer vision tasks like image captioning. This paper presents improvements to the soft attention model by combining a convolutional LSTM with a hierarchical system architecture to recognize action categories in videos. We call this model the Convolutional Hierarchical Attention Model (CHAM). The model applies a convolutional operation inside the LSTM cell and an attention map generation process to recognize actions. The hierarchical architecture of this model is able to explicitly reason on multi-granularities of action categories. The proposed architecture achieved improved results on three publicly available datasets: the UCF sports dataset, the Olympic sports dataset and the HMDB51 dataset.
- May 10 2017 cs.CV arXiv:1705.03148v1Visual data such as videos are often sampled from complex manifold. We propose leveraging the manifold structure to constrain the deep action feature learning, thereby minimizing the intra-class variations in the feature space and alleviating the over-fitting problem. Considering that manifold can be transferred, layer by layer, from the data domain to the deep features, the manifold priori is posed from the top layer into the back propagation learning procedure of convolutional neural network (CNN). The resulting algorithm --Spatio-Temporal Manifold Network-- is solved with the efficient Alternating Direction Method of Multipliers and Backward Propagation (ADMM-BP). We theoretically show that STMN recasts the problem as projection over the manifold via an embedding method. The proposed approach is evaluated on two benchmark datasets, showing significant improvements to the baselines.
- May 09 2017 cs.SY arXiv:1705.02421v1For its high coefficient of performance and zero local emissions, the heat pump (HP) has recently become popular in North Europe and China. However, the integration of HPs may aggravate the daily peak-valley gap in distribution networks significantly.
- May 04 2017 cs.CV arXiv:1705.01450v2Steerable properties dominate the design of traditional filters, e.g., Gabor filters, and endow features the capability of dealing with spatial transformations. However, such excellent properties have not been well explored in the popular deep convolutional neural networks (DCNNs). In this paper, we propose a new deep model, termed Gabor Convolutional Networks (GCNs or Gabor CNNs), which incorporates Gabor filters into DCNNs to enhance the resistance of deep learned features to the orientation and scale changes. By only manipulating the basic element of DCNNs based on Gabor filters, i.e., the convolution operator, GCNs can be easily implemented and are compatible with any popular deep learning architecture. Experimental results demonstrate the super capability of our algorithm in recognizing objects, where the scale and rotation changes occur frequently. The proposed GCNs have much fewer learnable network parameters, and thus is easier to train with an end-to-end pipeline. To encourage further developments, the source code is released at Github.
- Voltage control plays an important role in the operation of electricity distribution networks, especially with high penetration of distributed energy resources. These resources introduces significant and fast varying uncertainties. In this paper, we focus on reactive power compensation to control voltage in the presence of uncertainties. We adopt a probabilistic approach that accounts for arbitrary correlations between renewable resources at each of the buses and we use the linearized DistFlow equations to model the distribution network. We then show that this optimization problem is convex for a wide variety of probabilistic distributions. Compared to conventional per-bus chance constraints, our formulation is more robust to uncertainty and more computationally tractable. We illustrate the results using standard IEEE distribution test feeders.
- Apr 28 2017 cs.CL arXiv:1704.08430v1Neural machine translation (NMT) heavily relies on an attention network to produce a context vector for each target word prediction. In practice, we find that context vectors for different target words are quite similar to one another and therefore are insufficient in discriminatively predicting target words. The reason for this might be that context vectors produced by the vanilla attention network are just a weighted sum of source representations that are invariant to decoder states. In this paper, we propose a novel GRU-gated attention model (GAtt) for NMT which enhances the degree of discrimination of context vectors by enabling source representations to be sensitive to the partial translation generated by the decoder. GAtt uses a gated recurrent unit (GRU) to combine two types of information: treating a source annotation vector originally produced by the bidirectional encoder as the history state while the corresponding previous decoder state as the input to the GRU. The GRU-combined information forms a new source annotation vector. In this way, we can obtain translation-sensitive source representations which are then feed into the attention network to generate discriminative context vectors. We further propose a variant that regards a source annotation vector as the current input while the previous decoder state as the history. Experiments on NIST Chinese-English translation tasks show that both GAtt-based models achieve significant improvements over the vanilla attentionbased NMT. Further analyses on attention weights and context vectors demonstrate the effectiveness of GAtt in improving the discrimination power of representations and handling the challenging issue of over-translation.
- 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.
- Breast cancer is one of the leading causes of cancer death among women worldwide. In clinical routine, automatic breast ultrasound (BUS) image segmentation is very challenging and essential for cancer diagnosis and treatment planning. Many BUS segmentation approaches have been studied in the last two decades, and have been proved to be effective on private datasets. Currently, the advancement of BUS image segmentation seems to meet its bottleneck. The improvement of the performance is increasingly challenging, and only few new approaches were published in the last several years. It is the time to look at the field by reviewing previous approaches comprehensively and to investigate the future directions. In this paper, we study the basic ideas, theories, pros and cons of the approaches, group them into categories, and extensively review each category in depth by discussing the principles, application issues, and advantages/disadvantages.
- Catastrophic forgetting is a problem of neural networks that loses the information of the first task after training the second task. Here, we propose incremental moment matching (IMM) to resolve this problem. IMM incrementally matches the moment of the posterior distribution of neural networks, which is trained for the first and the second task, respectively. To make the search space of posterior parameter smooth, the IMM procedure is complemented by various transfer learning techniques including weight transfer, L2-norm of the old and the new parameter, and a variant of dropout with the old parameter. We analyze our approach on various datasets including the MNIST, CIFAR-10, Caltech-UCSD-Birds, and Lifelog datasets. Experimental results show that IMM achieves state-of-the-art performance in a variety of datasets and can balance the information between an old and a new network.
- A vital aspect in energy storage planning and operation is to accurately model its operational cost, which mainly comes from the battery cell degradation. Battery degradation can be viewed as a complex material fatigue process that based on stress cycles. Rainflow algorithm is a popular way for cycle identification in material fatigue process, and has been extensively used in battery degradation assessment. However, the rainflow algorithm does not have a closed form, which makes the major difficulty to include it in optimization. In this paper, we prove the rainflow cycle-based cost is convex. Convexity enables the proposed degradation model to be incorporated in different battery optimization problems and guarantees the solution quality. We provide a subgradient algorithm to solve the problem. A case study on PJM regulation market demonstrates the effectiveness of the proposed degradation model in maximizing the battery operating profits as well as extending its lifetime.
- Traditional image clustering methods take a two-step approach, feature learning and clustering, sequentially. However, recent research results demonstrated that combining the separated phases in a unified framework and training them jointly can achieve a better performance. In this paper, we first introduce fully convolutional auto-encoders for image feature learning and then propose a unified clustering framework to learn image representations and cluster centers jointly based on a fully convolutional auto-encoder and soft $k$-means scores. At initial stages of the learning procedure, the representations extracted from the auto-encoder may not be very discriminative for latter clustering. We address this issue by adopting a boosted discriminative distribution, where high score assignments are highlighted and low score ones are de-emphasized. With the gradually boosted discrimination, clustering assignment scores are discriminated and cluster purities are enlarged. Experiments on several vision benchmark datasets show that our methods can achieve a state-of-the-art performance.
- Deep learning models (DLMs) are state-of-the-art techniques in speech recognition. However, training good DLMs can be time consuming especially for production-size models and corpora. Although several parallel training algorithms have been proposed to improve training efficiency, there is no clear guidance on which one to choose for the task in hand due to lack of systematic and fair comparison among them. In this paper we aim at filling this gap by comparing four popular parallel training algorithms in speech recognition, namely asynchronous stochastic gradient descent (ASGD), blockwise model-update filtering (BMUF), bulk synchronous parallel (BSP) and elastic averaging stochastic gradient descent (EASGD), on 1000-hour LibriSpeech corpora using feed-forward deep neural networks (DNNs) and convolutional, long short-term memory, DNNs (CLDNNs). Based on our experiments, we recommend using BMUF as the top choice to train acoustic models since it is most stable, scales well with number of GPUs, can achieve reproducible results, and in many cases even outperforms single-GPU SGD. ASGD can be used as a substitute in some cases.
- Mar 16 2017 cs.CV arXiv:1703.05243v2Given the progress in image recognition with recent data driven paradigms, it's still expensive to manually label a large training data to fit a convolutional neural network (CNN) model. This paper proposes a hybrid supervised-unsupervised method combining a pre-trained AlexNet with Latent Dirichlet Allocation (LDA) to extract image topics from both an unlabeled life-logging dataset and the COCO dataset. We generate the bag-of-words representations of an egocentric dataset from the softmax layer of AlexNet and use LDA to visualize the subject's living genre with duplicated images. We use a subset of COCO on 4 categories as ground truth, and define consistent rate to quantitatively analyze the performance of the method, it achieves 84% for consistent rate on average comparing to 18.75% from a raw CNN model. The method is capable of detecting false labels and multi-labels from COCO dataset. For scalability test, parallelization experiments are conducted with Harp-LDA on a Intel Knights Landing cluster: to extract 1,000 topic assignments for 241,035 COCO images, it takes 10 minutes with 60 threads.
- Mar 14 2017 cs.LG arXiv:1703.04318v1Advances in Machine Learning (ML) have led to its adoption as an integral component in many applications, including banking, medical diagnosis, and driverless cars. To further broaden the use of ML models, cloud-based services offered by Microsoft, Amazon, Google, and others have developed ML-as-a-service tools as black-box systems. However, ML classifiers are vulnerable to adversarial examples: inputs that are maliciously modified can cause the classifier to provide adversary-desired outputs. Moreover, it is known that adversarial examples generated on one classifier are likely to cause another classifier to make the same mistake, even if the classifiers have different architectures or are trained on disjoint datasets. This property, which is known as transferability, opens up the possibility of attacking black-box systems by generating adversarial examples on a substitute classifier and transferring the examples to the target classifier. Therefore, the key to protect black-box learning systems against the adversarial examples is to block their transferability. To this end, we propose a training method that, as the input is more perturbed, the classifier smoothly outputs lower confidence on the original label and instead predicts that the input is "invalid". In essence, we augment the output class set with a NULL label and train the classifier to reject the adversarial examples by classifying them as NULL. In experiments, we apply a wide range of attacks based on adversarial examples on the black-box systems. We show that a classifier trained with the proposed method effectively resists against the adversarial examples, while maintaining the accuracy on clean data.
- Mar 14 2017 cs.CV arXiv:1703.04096v2Interpretability of deep neural networks (DNNs) is essential since it enables users to understand the overall strengths and weaknesses of the models, conveys an understanding of how the models will behave in the future, and how to diagnose and correct potential problems. However, it is challenging to reason about what a DNN actually does due to its opaque or black-box nature. To address this issue, we propose a novel technique to improve the interpretability of DNNs by leveraging the rich semantic information embedded in human descriptions. By concentrating on the video captioning task, we first extract a set of semantically meaningful topics from the human descriptions that cover a wide range of visual concepts, and integrate them into the model with an interpretive loss. We then propose a prediction difference maximization algorithm to interpret the learned features of each neuron. Experimental results demonstrate its effectiveness in video captioning using the interpretable features, which can also be transferred to video action recognition. By clearly understanding the learned features, users can easily revise false predictions via a human-in-the-loop procedure.
- Mar 14 2017 cs.AI arXiv:1703.03933v1Recently, reinforcement learning has been successfully applied to the logical game of Go, various Atari games, and even a 3D game, Labyrinth, though it continues to have problems in sparse reward settings. It is difficult to explore, but also difficult to exploit, a small number of successes when learning policy. To solve this issue, the subgoal and option framework have been proposed. However, discovering subgoals online is too expensive to be used to learn options in large state spaces. We propose Micro-objective learning (MOL) to solve this problem. The main idea is to estimate how important a state is while training and to give an additional reward proportional to its importance. We evaluated our algorithm in two Atari games: Montezuma's Revenge and Seaquest. With three experiments to each game, MOL significantly improved the baseline scores. Especially in Montezuma's Revenge, MOL achieved two times better results than the previous state-of-the-art model.
- Generative Adversarial Nets (GANs) have shown promise in image generation and semi-supervised learning (SSL). However, existing GANs in SSL have two problems: (1) the generator and the discriminator (i.e. the classifier) may not be optimal at the same time; and (2) the generator cannot control the semantics of the generated samples. The problems essentially arise from the two-player formulation, where a single discriminator shares incompatible roles of identifying fake samples and predicting labels and it only estimates the data without considering the labels. To address the problems, we present triple generative adversarial net (Triple-GAN), which consists of three players---a generator, a discriminator and a classifier. The generator and the classifier characterize the conditional distributions between images and labels, and the discriminator solely focuses on identifying fake image-label pairs. We design compatible utilities to ensure that the distributions characterized by the classifier and the generator both converge to the data distribution. Our results on various datasets demonstrate that Triple-GAN as a unified model can simultaneously (1) achieve the state-of-the-art classification results among deep generative models, and (2) disentangle the classes and styles of the input and transfer smoothly in the data space via interpolation in the latent space class-conditionally.
- We consider using a battery storage system simultaneously for peak shaving and frequency regulation through a joint optimization framework which captures battery degradation, operational constraints and uncertainties in customer load and regulation signals. Under this framework, using real data we show the electricity bill of users can be reduced by up to 15\%. Furthermore, we demonstrate that the saving from joint optimization is often larger than the sum of the optimal savings when the battery is used for the two individual applications. A simple threshold real-time algorithm is proposed and achieves this super-linear gain. Compared to prior works that focused on using battery storage systems for single applications, our results suggest that batteries can achieve much larger economic benefits than previously thought if they jointly provide multiple services.
- Social media platforms provide an environment where people can freely engage in discussions. Unfortunately, they also enable several problems, such as online harassment. Recently, Google and Jigsaw started a project called Perspective, which uses machine learning to automatically detect toxic language. A demonstration website has been also launched, which allows anyone to type a phrase in the interface and instantaneously see the toxicity score [1]. In this paper, we propose an attack on the Perspective toxic detection system based on the adversarial examples. We show that an adversary can subtly modify a highly toxic phrase in a way that the system assigns significantly lower toxicity score to it. We apply the attack on the sample phrases provided in the Perspective website and show that we can consistently reduce the toxicity scores to the level of the non-toxic phrases. The existence of such adversarial examples is very harmful for toxic detection systems and seriously undermines their usability.
- Feb 22 2017 cs.CY arXiv:1702.06156v1In this paper we explore city-level traffic and parking data to determine how much cruising for curbside parking contributes to overall traffic congestion. To this end, we describe a new kind of queueing network and present a data-informed model based on this new queuing network. We leverage the data-informed model in developing and validating a simulation tool. In addition, we utilize curbside parking and arterial traffic volume data to produce an estimate of the proportion of traffic searching for parking along high occupancy arterials. Somewhat surprisingly, we find that while percentage increase in travel time to through traffic vehicles depends on time of day, it does not appear to depend on high volumes of through traffic. Moreover, we show that the probability of a block-face being full is a much more viable metric for directly controlling congestion than average occupancy rate, typically used by municipalities.
- We study a demand response problem from utility (also referred to as operator)'s perspective with realistic settings, in which the utility faces uncertainty and limited communication. Specifically, the utility does not know the cost function of consumers and cannot have multiple rounds of information exchange with consumers. We formulate an optimization problem for the utility to minimize its operational cost considering time-varying demand response targets and responses of consumers. We develop a joint online learning and pricing algorithm. In each time slot, the utility sends out a price signal to all consumers and estimates the cost functions of consumers based on their noisy responses. We measure the performance of our algorithm using regret analysis and show that our online algorithm achieves logarithmic regret with respect to the operating horizon. In addition, our algorithm employs linear regression to estimate the aggregate response of consumers, making it easy to implement in practice. Simulation experiments validate the theoretic results and show that the performance gap between our algorithm and the offline optimality decays quickly.
- In real-world, our DNA is unique but many people share names. This phenomenon often causes erroneous aggregation of documents of multiple persons who are namesake of one another. Such mistakes deteriorate the performance of document retrieval, web search, and more seriously, cause improper attribution of credit or blame in digital forensic. To resolve this issue, the name disambiguation task is designed which aims to partition the documents associated with a name reference such that each partition contains documents pertaining to a unique real-life person. Existing solutions to this task substantially rely on feature engineering, such as biographical feature extraction, or construction of auxiliary features from Wikipedia. However, for many scenarios, such features may be costly to obtain or unavailable due to the risk of privacy violation. In this work, we propose a novel name disambiguation method. Our proposed method is non-intrusive of privacy because instead of using attributes pertaining to a real-life person, our method leverages only relational data in the form of anonymized graphs. In the methodological aspect, the proposed method uses a novel representation learning model to embed each document in a low dimensional vector space where name disambiguation can be solved by a hierarchical agglomerative clustering algorithm. Our experimental results demonstrate that the proposed method is significantly better than the existing name disambiguation methods working in a similar setting.
- Feb 08 2017 cs.CG arXiv:1702.01836v1We study approximation algorithms for the following geometric version of the maximum coverage problem: Let $\mathcal{P}$ be a set of $n$ weighted points in the plane. Let $D$ represent a planar object, such as a rectangle, or a disk. We want to place $m$ copies of $D$ such that the sum of the weights of the points in $\mathcal{P}$ covered by these copies is maximized. For any fixed $\varepsilon>0$, we present efficient approximation schemes that can find a $(1-\varepsilon)$-approximation to the optimal solution. In particular, for $m=1$ and for the special case where $D$ is a rectangle, our algorithm runs in time $O(n\log (\frac{1}{\varepsilon}))$, improving on the previous result. For $m>1$ and the rectangular case, our algorithm runs in $O(\frac{n}{\varepsilon}\log (\frac{1}{\varepsilon})+\frac{m}{\varepsilon}\log m +m(\frac{1}{\varepsilon})^{O(\min(\sqrt{m},\frac{1}{\varepsilon}))})$ time. For a more general class of shapes (including disks, polygons with $O(1)$ edges), our algorithm runs in $O(n(\frac{1}{\varepsilon})^{O(1)}+\frac{m}{\epsilon}\log m + m(\frac{1}{\varepsilon})^{O(\min(m,\frac{1}{\varepsilon^2}))})$ time.
- Jan 06 2017 cs.MA arXiv:1701.01289v1Multi-agent systems (MAS) is able to characterize the behavior of individual agent and the interaction between agents. Thus, it motivates us to leverage the distributed constraint optimization problem (DCOP), a framework of modeling MAS, to solve the user association problem in heterogeneous networks (HetNets). Two issues we have to consider when we take DCOP into the application of HetNet including: (i) How to set up an effective model by DCOP taking account of the negtive impact of the increment of users on the modeling process (ii) Which kind of algorithms is more suitable to balance the time consumption and the quality of soltuion. Aiming to overcome these issues, we firstly come up with an ECAV-$\eta$ (Each Connection As Variable) model in which a parameter $\eta$ with an adequate assignment ($\eta=3$ in this paper) is able to control the scale of the model. After that, a Markov chain (MC) based algorithm is proposed on the basis of log-sum-exp function. Experimental results show that the solution obtained by DCOP framework is better than the one obtained by the Max-SINR algorithm. Comparing with the Lagrange dual decomposition based method (LDD), the solution performance has been improved since there is no need to transform original problem into a satisfied one. In addition, it is also apparent that the DCOP based method has better robustness than LDD when the number of users increases but the available resource at base stations are limited.
- Based on the observation that the correlation between observed traffic at two measurement points or traffic stations may be time-varying, attributable to the time-varying speed which subsequently causes variations in the time required to travel between the two points, in this paper, we develop a modified Space-Time Autoregressive Integrated Moving Average (STARIMA) model with time-varying lags for short-term traffic flow prediction. Particularly, the temporal lags in the modified STARIMA change with the time-varying speed at different time of the day or equivalently change with the (time-varying) time required to travel between two measurement points. Firstly, a technique is developed to evaluate the temporal lag in the STARIMA model, where the temporal lag is formulated as a function of the spatial lag (spatial distance) and the average speed. Secondly, an unsupervised classification algorithm based on ISODATA algorithm is designed to classify different time periods of the day according to the variation of the speed. The classification helps to determine the appropriate time lag to use in the STARIMA model. Finally, a STARIMA-based model with time-varying lags is developed for short-term traffic prediction. Experimental results using real traffic data show that the developed STARIMA-based model with time-varying lags has superior accuracy compared with its counterpart developed using the traditional cross-correlation function and without employing time-varying lags.
- Dec 22 2016 cs.LG arXiv:1612.07146v2Deep neural networks have shown promise in collaborative filtering (CF). However, existing neural approaches are either user-based or item-based, which cannot leverage all the underlying information explicitly. We propose CF-UIcA, a neural co-autoregressive model for CF tasks, which exploits the structural correlation in the domains of both users and items. The co-autoregression allows extra desired properties to be incorporated for different tasks. Furthermore, we develop an efficient stochastic learning algorithm to handle large scale datasets. We evaluate CF-UIcA on two popular benchmarks: MovieLens 1M and Netflix, and achieve state-of-the-art performance in both rating prediction and top-N recommendation tasks, which demonstrates the effectiveness of CF-UIcA.
- Dec 19 2016 cs.CV arXiv:1612.05365v1Kernelized Correlation Filter (KCF) is one of the state-of-the-art object trackers. However, it does not reasonably model the distribution of correlation response during tracking process, which might cause the drifting problem, especially when targets undergo significant appearance changes due to occlusion, camera shaking, and/or deformation. In this paper, we propose an Output Constraint Transfer (OCT) method that by modeling the distribution of correlation response in a Bayesian optimization framework is able to mitigate the drifting problem. OCT builds upon the reasonable assumption that the correlation response to the target image follows a Gaussian distribution, which we exploit to select training samples and reduce model uncertainty. OCT is rooted in a new theory which transfers data distribution to a constraint of the optimized variable, leading to an efficient framework to calculate correlation filters. Extensive experiments on a commonly used tracking benchmark show that the proposed method significantly improves KCF, and achieves better performance than other state-of-the-art trackers. To encourage further developments, the source code is made available https://github.com/bczhangbczhang/OCT-KCF.
- Nov 24 2016 cs.CV arXiv:1611.07544v1In this paper, a self-learning approach is proposed towards solving scene-specific pedestrian detection problem without any human' annotation involved. The self-learning approach is deployed as progressive steps of object discovery, object enforcement, and label propagation. In the learning procedure, object locations in each frame are treated as latent variables that are solved with a progressive latent model (PLM). Compared with conventional latent models, the proposed PLM incorporates a spatial regularization term to reduce ambiguities in object proposals and to enforce object localization, and also a graph-based label propagation to discover harder instances in adjacent frames. With the difference of convex (DC) objective functions, PLM can be efficiently optimized with a concave-convex programming and thus guaranteeing the stability of self-learning. Extensive experiments demonstrate that even without annotation the proposed self-learning approach outperforms weakly supervised learning approaches, while achieving comparable performance with transfer learning and fully supervised approaches.
- Deep generative models (DGMs) are effective on learning multilayered representations of complex data and performing inference of input data by exploring the generative ability. However, it is relatively insufficient to empower the discriminative ability of DGMs on making accurate predictions. This paper presents max-margin deep generative models (mmDGMs) and a class-conditional variant (mmDCGMs), which explore the strongly discriminative principle of max-margin learning to improve the predictive performance of DGMs in both supervised and semi-supervised learning, while retaining the generative capability. In semi-supervised learning, we use the predictions of a max-margin classifier as the missing labels instead of performing full posterior inference for efficiency; we also introduce additional max-margin and label-balance regularization terms of unlabeled data for effectiveness. We develop an efficient doubly stochastic subgradient algorithm for the piecewise linear objectives in different settings. Empirical results on various datasets demonstrate that: (1) max-margin learning can significantly improve the prediction performance of DGMs and meanwhile retain the generative ability; (2) in supervised learning, mmDGMs are competitive to the best fully discriminative networks when employing convolutional neural networks as the generative and recognition models; and (3) in semi-supervised learning, mmDCGMs can perform efficient inference and achieve state-of-the-art classification results on several benchmarks.
- The $k$-means clustering algorithm is popular but has the following main drawbacks: 1) the number of clusters, $k$, needs to be provided by the user in advance, 2) it can easily reach local minima with randomly selected initial centers, 3) it is sensitive to outliers, and 4) it can only deal with well separated hyperspherical clusters. In this paper, we propose a Local Density Peaks Searching (LDPS) initialization framework to address these issues. The LDPS framework includes two basic components: one of them is the local density that characterizes the density distribution of a data set, and the other is the local distinctiveness index (LDI) which we introduce to characterize how distinctive a data point is compared with its neighbors. Based on these two components, we search for the local density peaks which are characterized with high local densities and high LDIs to deal with 1) and 2). Moreover, we detect outliers characterized with low local densities but high LDIs, and exclude them out before clustering begins. Finally, we apply the LDPS initialization framework to $k$-medoids, which is a variant of $k$-means and chooses data samples as centers, with diverse similarity measures other than the Euclidean distance to fix the last drawback of $k$-means. Combining the LDPS initialization framework with $k$-means and $k$-medoids, we obtain two novel clustering methods called LDPS-means and LDPS-medoids, respectively. Experiments on synthetic data sets verify the effectiveness of the proposed methods, especially when the ground truth of the cluster number $k$ is large. Further, experiments on several real world data sets, Handwritten Pendigits, Coil-20, Coil-100 and Olivetti Face Database, illustrate that our methods give a superior performance than the analogous approaches on both estimating $k$ and unsupervised object categorization.
- The GANs are generative models whose random samples realistically reflect natural images. It also can generate samples with specific attributes by concatenating a condition vector into the input, yet research on this field is not well studied. We propose novel methods of conditioning generative adversarial networks (GANs) that achieve state-of-the-art results on MNIST and CIFAR-10. We mainly introduce two models: an information retrieving model that extracts conditional information from the samples, and a spatial bilinear pooling model that forms bilinear features derived from the spatial cross product of an image and a condition vector. These methods significantly enhance log-likelihood of test data under the conditional distributions compared to the methods of concatenation.
- Routing in NDN networks must scale in terms of forwarding table size and routing protocol overhead. Hyperbolic routing (HR) presents a potential solution to address the routing scalability problem, because it does not use traditional forwarding tables or exchange routing updates upon changes in network topologies. Although HR has the drawbacks of producing sub-optimal routes or local minima for some destinations, these issues can be mitigated by NDN's intelligent data forwarding plane. However, HR's viability still depends on both the quality of the routes HR provides and the overhead incurred at the forwarding plane due to HR's sub-optimal behavior. We designed a new forwarding strategy called Adaptive Smoothed RTT-based Forwarding (ASF) to mitigate HR's sub-optimal path selection. This paper describes our experimental investigation into the packet delivery delay and overhead under HR as compared with Named-Data Link State Routing (NLSR), which calculates shortest paths. We run emulation experiments using various topologies with different failure scenarios, probing intervals, and maximum number of next hops for a name prefix. Our results show that HR's delay stretch has a median close to 1 and a 95th-percentile around or below 2, which does not grow with the network size. HR's message overhead in dynamic topologies is nearly independent of the network size, while NLSR's overhead grows polynomially at least. These results suggest that HR offers a more scalable routing solution with little impact on the optimality of routing paths.
- Bilinear models provide rich representations compared with linear models. They have been applied in various visual tasks, such as object recognition, segmentation, and visual question-answering, to get state-of-the-art performances taking advantage of the expanded representations. However, bilinear representations tend to be high-dimensional, limiting the applicability to computationally complex tasks. We propose low-rank bilinear pooling using Hadamard product for an efficient attention mechanism of multimodal learning. We show that our model outperforms compact bilinear pooling in visual question-answering tasks with the state-of-the-art results on the VQA dataset, having a better parsimonious property.
- Demand response is designed to motivate electricity customers to modify their loads at critical time periods. The accurate estimation of impact of demand response signals to customers' consumption is central to any successful program. In practice, learning these response is nontrivial because operators can only send a limited number of signals. In addition, customer behavior also depends on a large number of exogenous covariates. These two features lead to a high dimensional inference problem with limited number of observations. In this paper, we formulate this problem by using a multivariate linear model and adopt an experimental design approach to estimate the impact of demand response signals. We show that randomized assignment, which is widely used to estimate the average treatment effect, is not efficient in reducing the variance of the estimator when a large number of covariates is present. In contrast, we present a tractable algorithm that strategically assigns demand response signals to customers. This algorithm achieves the optimal reduction in estimation variance, independent of the number of covariates. The results are validated from simulations on synthetic data.
- Sep 29 2016 cs.CV arXiv:1609.08740v1Hashing method maps similar data to binary hashcodes with smaller hamming distance, and it has received a broad attention due to its low storage cost and fast retrieval speed. However, the existing limitations make the present algorithms difficult to deal with large-scale datasets: (1) discrete constraints are involved in the learning of the hash function; (2) pairwise or triplet similarity is adopted to generate efficient hashcodes, resulting both time and space complexity are greater than O(n^2). To address these issues, we propose a novel discrete supervised hash learning framework which can be scalable to large-scale datasets. First, the discrete learning procedure is decomposed into a binary classifier learning scheme and binary codes learning scheme, which makes the learning procedure more efficient. Second, we adopt the Asymmetric Low-rank Matrix Factorization and propose the Fast Clustering-based Batch Coordinate Descent method, such that the time and space complexity is reduced to O(n). The proposed framework also provides a flexible paradigm to incorporate with arbitrary hash function, including deep neural networks and kernel methods. Experiments on large-scale datasets demonstrate that the proposed method is superior or comparable with state-of-the-art hashing algorithms.
- Personal robots are expected to interact with the user by recognizing the user's face. However, in most of the service robot applications, the user needs to move himself/herself to allow the robot to see him/her face to face. To overcome such limitations, a method for estimating human body orientation is required. Previous studies used various components such as feature extractors and classification models to classify the orientation which resulted in low performance. For a more robust and accurate approach, we propose the light weight convolutional neural networks, an end to end system, for estimating human body orientation. Our body orientation estimation model achieved 81.58% and 94% accuracy with the benchmark dataset and our own dataset respectively. The proposed method can be used in a wide range of service robot applications which depend on the ability to estimate human body orientation. To show its usefulness in service robot applications, we designed a simple robot application which allows the robot to move towards the user's frontal plane. With this, we demonstrated an improved face detection rate.
- Sep 02 2016 cs.CV arXiv:1609.00153v2Traditional feature encoding scheme (e.g., Fisher vector) with local descriptors (e.g., SIFT) and recent convolutional neural networks (CNNs) are two classes of successful methods for image recognition. In this paper, we propose a hybrid representation, which leverages the discriminative capacity of CNNs and the simplicity of descriptor encoding schema for image recognition, with a focus on scene recognition. To this end, we make three main contributions from the following aspects. First, we propose a patch-level and end-to-end architecture to model the appearance of local patches, called \em PatchNet. PatchNet is essentially a customized network trained in a weakly supervised manner, which uses the image-level supervision to guide the patch-level feature extraction. Second, we present a hybrid visual representation, called \em VSAD, by utilizing the robust feature representations of PatchNet to describe local patches and exploiting the semantic probabilities of PatchNet to aggregate these local patches into a global representation. Third, based on the proposed VSAD representation, we propose a new state-of-the-art scene recognition approach, which achieves an excellent performance on two standard benchmarks: MIT Indoor67 (86.2\%) and SUN397 (73.0\%).
- We consider the setting where a collection of time series, modeled as random processes, evolve in a causal manner, and one is interested in learning the graph governing the relationships of these processes. A special case of wide interest and applicability is the setting where the noise is Gaussian and relationships are Markov and linear. We study this setting with two additional features: firstly, each random process has a hidden (latent) state, which we use to model the internal memory possessed by the variables (similar to hidden Markov models). Secondly, each variable can depend on its latent memory state through a random lag (rather than a fixed lag), thus modeling memory recall with differing lags at distinct times. Under this setting, we develop an estimator and prove that under a genericity assumption, the parameters of the model can be learned consistently. We also propose a practical adaption of this estimator, which demonstrates significant performance gains in both synthetic and real-world datasets.
- Aug 18 2016 cs.SY arXiv:1608.04879v1In distribution networks, there are slow controlling devices and fast controlling devices for Volt-VAR regulation. These slow controlling devices, such as capacitors or voltage regulators, cannot be operated frequently and should be scheduled tens of minutes ahead (Hereafter named as slow control). Since of the uncertainties in predicting the load and distributed generation, the voltage violations cannot be eliminated by fast controlling devices with improper schedule of the slow controlling devices. In this paper we propose dual time-scale coordination for the Volt-VAR control scheme, corresponding to slow and fast control. In the case of slow control, a robust voltage and reactive power optimization model is developed. This guarantees that subsequent fast controls can maintain the system's voltage security if the uncertain parameters vary within predefined limits. This nonconvex optimization problem is relaxed to a mix integer second order conic problem, and the dual form of its sub-problem is also derived. Then a column-and-constraint generation algorithm was used to solve the robust convexified model. A conventional deterministic optimization model can be used to determine the fast control mechanism. Numerical tests were conducted on a real distribution feeder in China, a balanced IEEE 69-bus and unbalanced 123-bus benchmark distribution networks. The simulation results show that solving the deterministic model is not always feasible and voltage violation may occur. The robust model was shown to be effective with respect to all possible scenarios in the uncertainty set, with little compromise in terms of network losses.
- Aug 09 2016 cs.CV arXiv:1608.02236v1Recently significant performance improvement in face detection was made possible by deeply trained convolutional networks. In this report, a novel approach for training state-of-the-art face detector is described. The key is to exploit the idea of hard negative mining and iteratively update the Faster R-CNN based face detector with the hard negatives harvested from a large set of background examples. We demonstrate that our face detector outperforms state-of-the-art detectors on the FDDB dataset, which is the de facto standard for evaluating face detection algorithms.