Showing papers on "Robustness (computer science) published in 2013"

ICDAR 2013 Robust Reading Competition

Abstract: This report presents the final results of the ICDAR 2013 Robust Reading Competition. The competition is structured in three Challenges addressing text extraction in different application domains, namely born-digital images, real scene images and real-scene videos. The Challenges are organised around specific tasks covering text localisation, text segmentation and word recognition. The competition took place in the first quarter of 2013, and received a total of 42 submissions over the different tasks offered. This report describes the datasets and ground truth specification, details the performance evaluation protocols used and presents the final results along with a brief summary of the participating methods.

Robust Face Landmark Estimation under Occlusion

Abstract: Human faces captured in real-world conditions present large variations in shape and occlusions due to differences in pose, expression, use of accessories such as sunglasses and hats and interactions with objects (e.g. food). Current face landmark estimation approaches struggle under such conditions since they fail to provide a principled way of handling outliers. We propose a novel method, called Robust Cascaded Pose Regression (RCPR) which reduces exposure to outliers by detecting occlusions explicitly and using robust shape-indexed features. We show that RCPR improves on previous landmark estimation methods on three popular face datasets (LFPW, LFW and HELEN). We further explore RCPR's performance by introducing a novel face dataset focused on occlusion, composed of 1,007 faces presenting a wide range of occlusion patterns. RCPR reduces failure cases by half on all four datasets, at the same time as it detects face occlusions with a 80/40% precision/recall.

An investigation of deep neural networks for noise robust speech recognition

Abstract: Recently, a new acoustic model based on deep neural networks (DNN) has been introduced. While the DNN has generated significant improvements over GMM-based systems on several tasks, there has been no evaluation of the robustness of such systems to environmental distortion. In this paper, we investigate the noise robustness of DNN-based acoustic models and find that they can match state-of-the-art performance on the Aurora 4 task without any explicit noise compensation. This performance can be further improved by incorporating information about the environment into DNN training using a new method called noise-aware training. When combined with the recently proposed dropout training technique, a 7.5% relative improvement over the previously best published result on this task is achieved using only a single decoding pass and no additional decoding complexity compared to a standard DNN.

Robust 1-Bit Compressive Sensing via Binary Stable Embeddings of Sparse Vectors

Abstract: The compressive sensing (CS) framework aims to ease the burden on analog-to-digital converters (ADCs) by reducing the sampling rate required to acquire and stably recover sparse signals. Practical ADCs not only sample but also quantize each measurement to a finite number of bits; moreover, there is an inverse relationship between the achievable sampling rate and the bit depth. In this paper, we investigate an alternative CS approach that shifts the emphasis from the sampling rate to the number of bits per measurement. In particular, we explore the extreme case of 1-bit CS measurements, which capture just their sign. Our results come in two flavors. First, we consider ideal reconstruction from noiseless 1-bit measurements and provide a lower bound on the best achievable reconstruction error. We also demonstrate that i.i.d. random Gaussian matrices provide measurement mappings that, with overwhelming probability, achieve nearly optimal error decay. Next, we consider reconstruction robustness to measurement errors and noise and introduce the binary e-stable embedding property, which characterizes the robustness of the measurement process to sign changes. We show that the same class of matrices that provide almost optimal noiseless performance also enable such a robust mapping. On the practical side, we introduce the binary iterative hard thresholding algorithm for signal reconstruction from 1-bit measurements that offers state-of-the-art performance.

Unified temporal and spatial calibration for multi-sensor systems

Abstract: In order to increase accuracy and robustness in state estimation for robotics, a growing number of applications rely on data from multiple complementary sensors. For the best performance in sensor fusion, these different sensors must be spatially and temporally registered with respect to each other. To this end, a number of approaches have been developed to estimate these system parameters in a two stage process, first estimating the time offset and subsequently solving for the spatial transformation between sensors. In this work, we present on a novel framework for jointly estimating the temporal offset between measurements of different sensors and their spatial displacements with respect to each other. The approach is enabled by continuous-time batch estimation and extends previous work by seamlessly incorporating time offsets within the rigorous theoretical framework of maximum likelihood estimation. Experimental results for a camera to inertial measurement unit (IMU) calibration prove the ability of this framework to accurately estimate time offsets up to a fraction of the smallest measurement period.

Fast Cost-Volume Filtering for Visual Correspondence and Beyond

Abstract: Many computer vision tasks can be formulated as labeling problems. The desired solution is often a spatially smooth labeling where label transitions are aligned with color edges of the input image. We show that such solutions can be efficiently achieved by smoothing the label costs with a very fast edge-preserving filter. In this paper, we propose a generic and simple framework comprising three steps: 1) constructing a cost volume, 2) fast cost volume filtering, and 3) Winner-Takes-All label selection. Our main contribution is to show that with such a simple framework state-of-the-art results can be achieved for several computer vision applications. In particular, we achieve 1) disparity maps in real time whose quality exceeds those of all other fast (local) approaches on the Middlebury stereo benchmark, and 2) optical flow fields which contain very fine structures as well as large displacements. To demonstrate robustness, the few parameters of our framework are set to nearly identical values for both applications. Also, competitive results for interactive image segmentation are presented. With this work, we hope to inspire other researchers to leverage this framework to other application areas.

Resilient Asymptotic Consensus in Robust Networks

Abstract: This paper addresses the problem of resilient in-network consensus in the presence of misbehaving nodes. Secure and fault-tolerant consensus algorithms typically assume knowledge of nonlocal information; however, this assumption is not suitable for large-scale dynamic networks. To remedy this, we focus on local strategies that provide resilience to faults and compromised nodes. We design a consensus protocol based on local information that is resilient to worst-case security breaches, assuming the compromised nodes have full knowledge of the network and the intentions of the other nodes. We provide necessary and sufficient conditions for the normal nodes to reach asymptotic consensus despite the influence of the misbehaving nodes under different threat assumptions. We show that traditional metrics such as connectivity are not adequate to characterize the behavior of such algorithms, and develop a novel graph-theoretic property referred to as network robustness. Network robustness formalizes the notion of redundancy of direct information exchange between subsets of nodes in the network, and is a fundamental property for analyzing the behavior of certain distributed algorithms that use only local information.

A robust and modular multi-sensor fusion approach applied to MAV navigation

Abstract: It has been long known that fusing information from multiple sensors for robot navigation results in increased robustness and accuracy. However, accurate calibration of the sensor ensemble prior to deployment in the field as well as coping with sensor outages, different measurement rates and delays, render multi-sensor fusion a challenge. As a result, most often, systems do not exploit all the sensor information available in exchange for simplicity. For example, on a mission requiring transition of the robot from indoors to outdoors, it is the norm to ignore the Global Positioning System (GPS) signals which become freely available once outdoors and instead, rely only on sensor feeds (e.g., vision and laser) continuously available throughout the mission. Naturally, this comes at the expense of robustness and accuracy in real deployment. This paper presents a generic framework, dubbed MultiSensor-Fusion Extended Kalman Filter (MSF-EKF), able to process delayed, relative and absolute measurements from a theoretically unlimited number of different sensors and sensor types, while allowing self-calibration of the sensor-suite online. The modularity of MSF-EKF allows seamless handling of additional/lost sensor signals during operation while employing a state buffering scheme augmented with Iterated EKF (IEKF) updates to allow for efficient re-linearization of the prediction to get near optimal linearization points for both absolute and relative state updates. We demonstrate our approach in outdoor navigation experiments using a Micro Aerial Vehicle (MAV) equipped with a GPS receiver as well as visual, inertial, and pressure sensors.

Distributed Optimal Power Flow for Smart Microgrids

Abstract: Optimal power flow (OPF) is considered for microgrids, with the objective of minimizing either the power distribution losses, or, the cost of power drawn from the substation and supplied by distributed generation (DG) units, while effecting voltage regulation. The microgrid is unbalanced, due to unequal loads in each phase and non-equilateral conductor spacings on the distribution lines. Similar to OPF formulations for balanced systems, the considered OPF problem is nonconvex. Nevertheless, a semidefinite programming (SDP) relaxation technique is advocated to obtain a convex problem solvable in polynomial-time complexity. Enticingly, numerical tests demonstrate the ability of the proposed method to attain the globally optimal solution of the original nonconvex OPF. To ensure scalability with respect to the number of nodes, robustness to isolated communication outages, and data privacy and integrity, the proposed SDP is solved in a distributed fashion by resorting to the alternating direction method of multipliers. The resulting algorithm entails iterative message-passing among groups of consumers and guarantees faster convergence compared to competing alternatives.

USAC: A Universal Framework for Random Sample Consensus

Abstract: A computational problem that arises frequently in computer vision is that of estimating the parameters of a model from data that have been contaminated by noise and outliers. More generally, any practical system that seeks to estimate quantities from noisy data measurements must have at its core some means of dealing with data contamination. The random sample consensus (RANSAC) algorithm is one of the most popular tools for robust estimation. Recent years have seen an explosion of activity in this area, leading to the development of a number of techniques that improve upon the efficiency and robustness of the basic RANSAC algorithm. In this paper, we present a comprehensive overview of recent research in RANSAC-based robust estimation by analyzing and comparing various approaches that have been explored over the years. We provide a common context for this analysis by introducing a new framework for robust estimation, which we call Universal RANSAC (USAC). USAC extends the simple hypothesize-and-verify structure of standard RANSAC to incorporate a number of important practical and computational considerations. In addition, we provide a general-purpose C++ software library that implements the USAC framework by leveraging state-of-the-art algorithms for the various modules. This implementation thus addresses many of the limitations of standard RANSAC within a single unified package. We benchmark the performance of the algorithm on a large collection of estimation problems. The implementation we provide can be used by researchers either as a stand-alone tool for robust estimation or as a benchmark for evaluating new techniques.

Real-Time 3D Reconstruction in Dynamic Scenes Using Point-Based Fusion

Abstract: Real-time or online 3D reconstruction has wide applicability and receives further interest due to availability of consumer depth cameras. Typical approaches use a moving sensor to accumulate depth measurements into a single model which is continuously refined. Designing such systems is an intricate balance between reconstruction quality, speed, spatial scale, and scene assumptions. Existing online methods either trade scale to achieve higher quality reconstructions of small objects/scenes. Or handle larger scenes by trading real-time performance and/or quality, or by limiting the bounds of the active reconstruction. Additionally, many systems assume a static scene, and cannot robustly handle scene motion or reconstructions that evolve to reflect scene changes. We address these limitations with a new system for real-time dense reconstruction with equivalent quality to existing online methods, but with support for additional spatial scale and robustness in dynamic scenes. Our system is designed around a simple and flat point-Based representation, which directly works with the input acquired from range/depth sensors, without the overhead of converting between representations. The use of points enables speed and memory efficiency, directly leveraging the standard graphics pipeline for all central operations, i.e., camera pose estimation, data association, outlier removal, fusion of depth maps into a single denoised model, and detection and update of dynamic objects. We conclude with qualitative and quantitative results that highlight robust tracking and high quality reconstructions of a diverse set of scenes at varying scales.

Attack Robustness and Centrality of Complex Networks

Abstract: Many complex systems can be described by networks, in which the constituent components are represented by vertices and the connections between the components are represented by edges between the corresponding vertices. A fundamental issue concerning complex networked systems is the robustness of the overall system to the failure of its constituent parts. Since the degree to which a networked system continues to function, as its component parts are degraded, typically depends on the integrity of the underlying network, the question of system robustness can be addressed by analyzing how the network structure changes as vertices are removed. Previous work has considered how the structure of complex networks change as vertices are removed uniformly at random, in decreasing order of their degree, or in decreasing order of their betweenness centrality. Here we extend these studies by investigating the effect on network structure of targeting vertices for removal based on a wider range of non-local measures of potential importance than simply degree or betweenness. We consider the effect of such targeted vertex removal on model networks with different degree distributions, clustering coefficients and assortativity coefficients, and for a variety of empirical networks.

Immunity to Device Variations in a Spiking Neural Network With Memristive Nanodevices

Abstract: Memristive nanodevices can feature a compact multilevel nonvolatile memory function, but are prone to device variability. We propose a novel neural network-based computing paradigm, which exploits their specific physics, and which has virtual immunity to their variability. Memristive devices are used as synapses in a spiking neural network performing unsupervised learning. They learn using a simplified and customized “spike timing dependent plasticity” rule. In the network, neurons’ threshold is adjusted following a homeostasis-type rule. We perform system level simulations with an experimentally verified model of the memristive devices’ behavior. They show, on the textbook case of character recognition, that performance can compare with traditional supervised networks of similar complexity. They also show that the system can retain functionality with extreme variations of various memristive devices’ parameters (a relative standard dispersion of more than 50% is tolerated on all device parameters), thanks to the robustness of the scheme, its unsupervised nature, and the capability of homeostasis. Additionally the network can adjust to stimuli presented with different coding schemes, is particularly robust to read disturb effects and does not require unrealistic control on the devices’ conductance. These results open the way for a novel design approach for ultraadaptive electronic systems.

Many objective robust decision making for complex environmental systems undergoing change

Abstract: This paper introduces many objective robust decision making (MORDM). MORDM combines concepts and methods from many objective evolutionary optimization and robust decision making (RDM), along with extensive use of interactive visual analytics, to facilitate the management of complex environmental systems. Many objective evolutionary search is used to generate alternatives for complex planning problems, enabling the discovery of the key tradeoffs among planning objectives. RDM then determines the robustness of planning alternatives to deeply uncertain future conditions and facilitates decision makers' selection of promising candidate solutions. MORDM tests each solution under the ensemble of future extreme states of the world (SOW). Interactive visual analytics are used to explore whether solutions of interest are robust to a wide range of plausible future conditions (i.e., assessment of their Pareto satisficing behavior in alternative SOW). Scenario discovery methods that use statistical data mining algorithms are then used to identify what assumptions and system conditions strongly influence the cost-effectiveness, efficiency, and reliability of the robust alternatives. The framework is demonstrated using a case study that examines a single city's water supply in the Lower Rio Grande Valley (LRGV) in Texas, USA. Results suggest that including robustness as a decision criterion can dramatically change the formulation of complex environmental management problems as well as the negotiated selection of candidate alternatives to implement. MORDM also allows decision makers to characterize the most important vulnerabilities for their systems, which should be the focus of ex post monitoring and identification of triggers for adaptive management.

Learning-based approach for online lane change intention prediction

Abstract: Predicting driver behavior is a key component for Advanced Driver Assistance Systems (ADAS). In this paper, a novel approach based on Support Vector Machine and Bayesian filtering is proposed for online lane change intention prediction. The approach uses the multiclass probabilistic outputs of the Support Vector Machine as an input to the Bayesian filter, and the output of the Bayesian filter is used for the final prediction of lane changes. A lane tracker integrated in a passenger vehicle is used for real-world data collection for the purpose of training and testing. Data from different drivers on different highways were used to evaluate the robustness of the approach. The results demonstrate that the proposed approach is able to predict driver intention to change lanes on average 1.3 seconds in advance, with a maximum prediction horizon of 3.29 seconds.

A Family of Nonparametric Density Estimation Algorithms

Abstract: A new methodology for density estimation is proposed. The methodology, which builds on the one developed by Tabak and Vanden-Eijnden, normalizes the data points through the composition of simple maps. The parameters of each map are determined through the maximization of a local quadratic approximation to the log-likelihood. Various candidates for the elementary maps of each step are proposed; criteria for choosing one includes robustness, computational simplicity, and good behavior in high-dimensional settings. A good choice is that of localized radial expansions, which depend on a single parameter: all the complexity of arbitrary, possibly convoluted probability densities can be built through the composition of such simple maps. © 2012 Wiley Periodicals, Inc.

A New Algorithm for Real-Time Multiple Open-Circuit Fault Diagnosis in Voltage-Fed PWM Motor Drives by the Reference Current Errors

Abstract: Three-phase inverters are currently utilized in an enormous variety of industrial applications, including variable-speed ac drives. However, due to their complexity and exposure to several stresses, they are prone to suffer critical failures. Accordingly, this paper presents a novel diagnostic algorithm that allows the real-time detection and localization of multiple power switch open-circuit faults in inverter-fed ac motor drives. The proposed method is quite simple and just requires the measured motor phase currents and their corresponding reference signals, already available from the main control system, therefore avoiding the use of additional sensors and hardware. Several experimental results using a vector-controlled permanent-magnet synchronous motor drive are presented, showing the diagnostic algorithm effectiveness, its relatively fast detection time, and its robustness against false alarms.

Adjustable Robust OPF With Renewable Energy Sources

Abstract: This paper presents an adjustable robust optimization approach to account for the uncertainty of renewable energy sources (RESs) in optimal power flow (OPF). It proposes an affinely adjustable robust OPF formulation where the base-point generation is calculated to serve the forecast load which is not balanced by RESs, and the generation control through participation factors ensures a feasible solution for all realizations of RES output within a prescribed uncertainty set. The adjustable robust OPF framework is solved using quadratic programming with successive constraint enforcement and can coordinate the computation of both the base-point generation and participation factors. Numerical results on standard test networks reveal a relatively small increase in the expected operational cost as the uncertainty level increases. In addition, solutions of networks that include both uncertain wind generation and Gaussian distributed demand are shown to have less cost and a higher level of robustness as compared to those from a recent robust scheduling method.

Robust stability and stabilization of uncertain linear positive systems via integral linear constraints : L-1-gain and L-infinity-gain characterization

Abstract: Copositive linear Lyapunov functions are used along with dissipativity theory for stability analysis and control of uncertain linear positive systems. Unlike usual results on linear systems, linear ...

Fault Detection for Modular Multilevel Converters Based on Sliding Mode Observer

Abstract: This letter presents a fault detection method for modular multilevel converters which is capable of locating a faulty semiconductor switching device in the circuit. The proposed fault detection method is based on a sliding mode observer (SMO) and a switching model of a half-bridge, the approach taken is to conjecture the location of fault, modify the SMO accordingly and then compare the observed and measured states to verify, or otherwise, the assumption. This technique requires no additional measurement elements and can easily be implemented in a DSP or microcontroller. The operation and robustness of the fault detection technique are confirmed by simulation results for the fault condition of a semiconductor switching device appearing as an open circuit.

Nonlinear-Disturbance-Observer-Based Robust Flight Control for Airbreathing Hypersonic Vehicles

Abstract: The work presented here is concerned with the robust flight control problem for the longitudinal dynamics of a generic airbreathing hypersonic vehicles (AHVs) under mismatched disturbances via a nonlinear-disturbance-observer-based control (NDOBC) method. Compared with other robust flight control method for AHV, the proposed method obtains not only promising robustness and disturbance rejection performance but also the property of nominal performance recovery. The merits of the proposed method are validated by simulation studies.

Online Robust PCA via Stochastic Optimization

Abstract: Robust PCA methods are typically based on batch optimization and have to load all the samples into memory during optimization. This prevents them from efficiently processing big data. In this paper, we develop an Online Robust PCA (OR-PCA) that processes one sample per time instance and hence its memory cost is independent of the number of samples, significantly enhancing the computation and storage efficiency. The proposed OR-PCA is based on stochastic optimization of an equivalent reformulation of the batch RPCA. Indeed, we show that OR-PCA provides a sequence of subspace estimations converging to the optimum of its batch counterpart and hence is provably robust to sparse corruption. Moreover, OR-PCA can naturally be applied for tracking dynamic subspace. Comprehensive simulations on subspace recovering and tracking demonstrate the robustness and efficiency advantages of the OR-PCA over online PCA and batch RPCA methods.

Robust Synchronization of Uncertain Linear Multi-Agent Systems

Abstract: This paper deals with robust synchronization of uncertain multi-agent networks. Given a network with for each of the agents identical nominal linear dynamics, we allow uncertainty in the form of additive perturbations of the transfer matrices of the nominal dynamics. The perturbations are assumed to be stable and bounded in H∞-norm by some a priori given desired tolerance. We derive state space formulas for observer based dynamic protocols that achieve synchronization for all perturbations bounded by this desired tolerance. It is shown that a protocol achieves robust synchronization if and only if each controller from a related finite set of feedback controllers robustly stabilizes a given, single linear system. Our protocols are expressed in terms of real symmetric solutions of certain algebraic Riccati equations and inequalities, and also involve weighting factors that depend on the eigenvalues of the graph Laplacian. For undirected network graphs we show that within the class of such dynamic protocols, a guaranteed achievable tolerance can be obtained that is proportional to the quotient of the second smallest and the largest eigenvalue of the Laplacian. We also extend our results to additive nonlinear perturbations with L2-gain bounded by a given tolerance.

The extreme vulnerability of interdependent spatially embedded networks

Abstract: Networks of networks are vulnerable: a failure in one sub-network can bring the rest crashing down. Previous simulations have suggested that randomly positioned networks might offer some limited robustness under certain circumstances. Analysis now shows, however, that real-world interdependent networks, where nodes are positioned according to geographical constraints, might not be so resilient.

Real-Time Articulated Hand Pose Estimation Using Semi-supervised Transductive Regression Forests

Abstract: This paper presents the first semi-supervised transductive algorithm for real-time articulated hand pose estimation. Noisy data and occlusions are the major challenges of articulated hand pose estimation. In addition, the discrepancies among realistic and synthetic pose data undermine the performances of existing approaches that use synthetic data extensively in training. We therefore propose the Semi-supervised Transductive Regression (STR) forest which learns the relationship between a small, sparsely labelled realistic dataset and a large synthetic dataset. We also design a novel data-driven, pseudo-kinematic technique to refine noisy or occluded joints. Our contributions include: (i) capturing the benefits of both realistic and synthetic data via transductive learning, (ii) showing accuracies can be improved by considering unlabelled data, and (iii) introducing a pseudo-kinematic technique to refine articulations efficiently. Experimental results show not only the promising performance of our method with respect to noise and occlusions, but also its superiority over state-of-the-arts in accuracy, robustness and speed.

Delay-Dependent Robust Load Frequency Control for Time Delay Power Systems

Abstract: The usage of communication channels introduces time delays into load frequency control (LFC) schemes. Those delays may degrade dynamic performance, and even cause instability, of a closed-loop LFC scheme. In this paper, a delay-dependent robust method is proposed for analysis/synthesis of a PID-type LFC scheme considering time delays. The effect of the disturbance on the controlled output is defined as a robust performance index (RPI) of the closed-loop system. At first, for a preset delay upper bound, controller gains are determined by minimizing the RPI. Secondly, calculation of the RPIs of the closed-loop system under different delays provides a new way to assess robustness against delays and estimate delay margins. Case studies are based on three-area LFC schemes under traditional and deregulated environments, respectively. The results show that the PID-type controller obtained can guarantee the tolerance for delays less than the preset upper bound and provide a bigger delay margin than the existing controllers do. Moreover, its robustness against load variations and parameter uncertainties is verified via simulation studies.

Hybrid BFOA-PSO algorithm for automatic generation control of linear and nonlinear interconnected power systems

Abstract: In the bacteria foraging optimization algorithm (BFAO), the chemotactic process is randomly set, imposing that the bacteria swarm together and keep a safe distance from each other. In hybrid bacteria foraging optimization algorithm and particle swarm optimization (hBFOA-PSO) algorithm the principle of swarming is introduced in the framework of BFAO. The hBFOA-PSO algorithm is based on the adjustment of each bacterium position according to the neighborhood environment. In this paper, the effectiveness of the hBFOA-PSO algorithm has been tested for automatic generation control (AGC) of an interconnected power system. A widely used linear model of two area non-reheat thermal system equipped with proportional-integral (PI) controller is considered initially for the design and analysis purpose. At first, a conventional integral time multiply absolute error (ITAE) based objective function is considered and the performance of hBFOA-PSO algorithm is compared with PSO, BFOA and GA. Further a modified objective function using ITAE, damping ratio of dominant eigenvalues and settling time with appropriate weight coefficients is proposed to increase the performance of the controller. Further, robustness analysis is carried out by varying the operating load condition and time constants of speed governor, turbine, tie-line power in the range of +50% to -50% as well as size and position of step load perturbation to demonstrate the robustness of the proposed hBFOA-PSO optimized PI controller. The proposed approach is also extended to a non-linear power system model by considering the effect of governor dead band non-linearity and the superiority of the proposed approach is shown by comparing the results of craziness based particle swarm optimization (CRAZYPSO) approach for the identical interconnected power system. Finally, the study is extended to a three area system considering both thermal and hydro units with different PI coefficients and comparison between ANFIS and proposed approach has been provided.