July 28, 2019

3175 words 15 mins read

Paper Group ANR 433

Learning Model Reparametrizations: Implicit Variational Inference by Fitting MCMC distributions. Light Source Estimation with Analytical Path-tracing. Bandit Regret Scaling with the Effective Loss Range. Towards Automated Network Mitigation Analysis (extended). Blocking Transferability of Adversarial Examples in Black-Box Learning Systems. When You …

Learning Model Reparametrizations: Implicit Variational Inference by Fitting MCMC distributions


Title	Learning Model Reparametrizations: Implicit Variational Inference by Fitting MCMC distributions
Authors	Michalis K. Titsias
Abstract	We introduce a new algorithm for approximate inference that combines reparametrization, Markov chain Monte Carlo and variational methods. We construct a very flexible implicit variational distribution synthesized by an arbitrary Markov chain Monte Carlo operation and a deterministic transformation that can be optimized using the reparametrization trick. Unlike current methods for implicit variational inference, our method avoids the computation of log density ratios and therefore it is easily applicable to arbitrary continuous and differentiable models. We demonstrate the proposed algorithm for fitting banana-shaped distributions and for training variational autoencoders.
Tasks
Published	2017-08-04
URL	http://arxiv.org/abs/1708.01529v1
PDF	http://arxiv.org/pdf/1708.01529v1.pdf
PWC	https://paperswithcode.com/paper/learning-model-reparametrizations-implicit
Repo
Framework

Light Source Estimation with Analytical Path-tracing


Title	Light Source Estimation with Analytical Path-tracing
Authors	Mike Kasper, Nima Keivan, Gabe Sibley, Christoffer Heckman
Abstract	We present a novel algorithm for light source estimation in scenes reconstructed with a RGB-D camera based on an analytically-derived formulation of path-tracing. Our algorithm traces the reconstructed scene with a custom path-tracer and computes the analytical derivatives of the light transport equation from principles in optics. These derivatives are then used to perform gradient descent, minimizing the photometric error between one or more captured reference images and renders of our current lighting estimation using an environment map parameterization for light sources. We show that our approach of modeling all light sources as points at infinity approximates lights located near the scene with surprising accuracy. Due to the analytical formulation of derivatives, optimization to the solution is considerably accelerated. We verify our algorithm using both real and synthetic data.
Tasks	Outdoor Light Source Estimation
Published	2017-01-15
URL	http://arxiv.org/abs/1701.04101v1
PDF	http://arxiv.org/pdf/1701.04101v1.pdf
PWC	https://paperswithcode.com/paper/light-source-estimation-with-analytical-path
Repo
Framework

Bandit Regret Scaling with the Effective Loss Range


Title	Bandit Regret Scaling with the Effective Loss Range
Authors	Nicolò Cesa-Bianchi, Ohad Shamir
Abstract	We study how the regret guarantees of nonstochastic multi-armed bandits can be improved, if the effective range of the losses in each round is small (e.g. the maximal difference between two losses in a given round). Despite a recent impossibility result, we show how this can be made possible under certain mild additional assumptions, such as availability of rough estimates of the losses, or advance knowledge of the loss of a single, possibly unspecified arm. Along the way, we develop a novel technique which might be of independent interest, to convert any multi-armed bandit algorithm with regret depending on the loss range, to an algorithm with regret depending only on the effective range, while avoiding predictably bad arms altogether.
Tasks	Multi-Armed Bandits
Published	2017-05-15
URL	https://arxiv.org/abs/1705.05091v3
PDF	https://arxiv.org/pdf/1705.05091v3.pdf
PWC	https://paperswithcode.com/paper/bandit-regret-scaling-with-the-effective-loss
Repo
Framework

Towards Automated Network Mitigation Analysis (extended)


Title	Towards Automated Network Mitigation Analysis (extended)
Authors	Patrick Speicher, Marcel Steinmetz, Jörg Hoffmann, Michael Backes, Robert Künnemann
Abstract	Penetration testing is a well-established practical concept for the identification of potentially exploitable security weaknesses and an important component of a security audit. Providing a holistic security assessment for networks consisting of several hundreds hosts is hardly feasible though without some sort of mechanization. Mitigation, prioritizing counter-measures subject to a given budget, currently lacks a solid theoretical understanding and is hence more art than science. In this work, we propose the first approach for conducting comprehensive what-if analyses in order to reason about mitigation in a conceptually well-founded manner. To evaluate and compare mitigation strategies, we use simulated penetration testing, i.e., automated attack-finding, based on a network model to which a subset of a given set of mitigation actions, e.g., changes to the network topology, system updates, configuration changes etc. is applied. Using Stackelberg planning, we determine optimal combinations that minimize the maximal attacker success (similar to a Stackelberg game), and thus provide a well-founded basis for a holistic mitigation strategy. We show that these Stackelberg planning models can largely be derived from network scan, public vulnerability databases and manual inspection with various degrees of automation and detail, and we simulate mitigation analysis on networks of different size and vulnerability.
Tasks
Published	2017-05-15
URL	http://arxiv.org/abs/1705.05088v2
PDF	http://arxiv.org/pdf/1705.05088v2.pdf
PWC	https://paperswithcode.com/paper/towards-automated-network-mitigation-analysis
Repo
Framework

Blocking Transferability of Adversarial Examples in Black-Box Learning Systems


Title	Blocking Transferability of Adversarial Examples in Black-Box Learning Systems
Authors	Hossein Hosseini, Yize Chen, Sreeram Kannan, Baosen Zhang, Radha Poovendran
Abstract	Advances 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.
Tasks	Medical Diagnosis
Published	2017-03-13
URL	http://arxiv.org/abs/1703.04318v1
PDF	http://arxiv.org/pdf/1703.04318v1.pdf
PWC	https://paperswithcode.com/paper/blocking-transferability-of-adversarial
Repo
Framework

When You Must Forget: beyond strong persistence when forgetting in answer set programming


Title	When You Must Forget: beyond strong persistence when forgetting in answer set programming
Authors	Ricardo Gonçalves, Matthias Knorr, João Leite, Stefan Woltran
Abstract	Among the myriad of desirable properties discussed in the context of forgetting in Answer Set Programming (ASP), strong persistence naturally captures its essence. Recently, it has been shown that it is not always possible to forget a set of atoms from a program while obeying this property, and a precise criterion regarding what can be forgotten has been presented, accompanied by a class of forgetting operators that return the correct result when forgetting is possible. However, it is an open question what to do when we have to forget a set of atoms, but cannot without violating this property. In this paper, we address this issue and investigate three natural alternatives to forget when forgetting without violating strong persistence is not possible, which turn out to correspond to the different possible relaxations of the characterization of strong persistence. Additionally, we discuss their preferable usage, shed light on the relation between forgetting and notions of relativized equivalence established earlier in the context of ASP, and present a detailed study on their computational complexity.
Tasks
Published	2017-07-17
URL	http://arxiv.org/abs/1707.05152v1
PDF	http://arxiv.org/pdf/1707.05152v1.pdf
PWC	https://paperswithcode.com/paper/when-you-must-forget-beyond-strong
Repo
Framework

Robust tracking of respiratory rate in high-dynamic range scenes using mobile thermal imaging


Title	Robust tracking of respiratory rate in high-dynamic range scenes using mobile thermal imaging
Authors	Youngjun Cho, Simon J. Julier, Nicolai Marquardt, Nadia Bianchi-Berthouze
Abstract	The ability to monitor respiratory rate is extremely important for medical treatment, healthcare and fitness sectors. In many situations, mobile methods, which allow users to undertake every day activities, are required. However, current monitoring systems can be obtrusive, requiring users to wear respiration belts or nasal probes. Recent advances in thermographic systems have shrunk their size, weight and cost, to the point where it is possible to create smart-phone based respiration rate monitoring devices that are not affected by lighting conditions. However, mobile thermal imaging is challenged in scenes with high thermal dynamic ranges. This challenge is further amplified by general problems such as motion artifacts and low spatial resolution, leading to unreliable breathing signals. In this paper, we propose a novel and robust approach for respiration tracking which compensates for the negative effects of variations in the ambient temperature and motion artifacts and can accurately extract breathing rates in highly dynamic thermal scenes. It has three main contributions. The first is a novel Optimal Quantization technique which adaptively constructs a color mapping of absolute temperature to improve segmentation, classification and tracking. The second is the Thermal Gradient Flow method that computes thermal gradient magnitude maps to enhance accuracy of the nostril region tracking. Finally, we introduce the Thermal Voxel method to increase the reliability of the captured respiration signals compared to the traditional averaging method. We demonstrate the extreme robustness of our system to track the nostril-region and measure the respiratory rate in high dynamic range scenes.
Tasks	Quantization
Published	2017-05-08
URL	http://arxiv.org/abs/1705.06628v2
PDF	http://arxiv.org/pdf/1705.06628v2.pdf
PWC	https://paperswithcode.com/paper/robust-tracking-of-respiratory-rate-in-high
Repo
Framework

Customized Nonlinear Bandits for Online Response Selection in Neural Conversation Models


Title	Customized Nonlinear Bandits for Online Response Selection in Neural Conversation Models
Authors	Bing Liu, Tong Yu, Ian Lane, Ole J. Mengshoel
Abstract	Dialog response selection is an important step towards natural response generation in conversational agents. Existing work on neural conversational models mainly focuses on offline supervised learning using a large set of context-response pairs. In this paper, we focus on online learning of response selection in retrieval-based dialog systems. We propose a contextual multi-armed bandit model with a nonlinear reward function that uses distributed representation of text for online response selection. A bidirectional LSTM is used to produce the distributed representations of dialog context and responses, which serve as the input to a contextual bandit. In learning the bandit, we propose a customized Thompson sampling method that is applied to a polynomial feature space in approximating the reward. Experimental results on the Ubuntu Dialogue Corpus demonstrate significant performance gains of the proposed method over conventional linear contextual bandits. Moreover, we report encouraging response selection performance of the proposed neural bandit model using the Recall@k metric for a small set of online training samples.
Tasks	Multi-Armed Bandits
Published	2017-11-22
URL	http://arxiv.org/abs/1711.08493v1
PDF	http://arxiv.org/pdf/1711.08493v1.pdf
PWC	https://paperswithcode.com/paper/customized-nonlinear-bandits-for-online
Repo
Framework

Feature Selection based on PCA and PSO for Multimodal Medical Image Fusion using DTCWT


Title	Feature Selection based on PCA and PSO for Multimodal Medical Image Fusion using DTCWT
Authors	Padmavathi K, Mahima Bhat, Maya V Karki
Abstract	Multimodal medical image fusion helps to increase efficiency in medical diagnosis. This paper presents multimodal medical image fusion by selecting relevant features using Principle Component Analysis (PCA) and Particle Swarm Optimization techniques (PSO). DTCWT is used for decomposition of the images into low and high frequency coefficients. Fusion rules such as combination of minimum, maximum and simple averaging are applied to approximate and detailed coefficients. The fused image is reconstructed by inverse DTCWT. Performance metrics are evaluated and it shows that DTCWT-PCA performs better than DTCWT-PSO in terms of Structural Similarity Index Measure (SSIM) and Cross Correlation (CC). Computation time and feature vector size is reduced in DTCWT-PCA compared to DTCWT-PSO for feature selection which proves robustness and storage capacity.
Tasks	Feature Selection, Medical Diagnosis
Published	2017-01-31
URL	http://arxiv.org/abs/1701.08918v1
PDF	http://arxiv.org/pdf/1701.08918v1.pdf
PWC	https://paperswithcode.com/paper/feature-selection-based-on-pca-and-pso-for
Repo
Framework

Conditional Markov Chain Search for the Simple Plant Location Problem improves upper bounds on twelve Körkel-Ghosh instances


Title	Conditional Markov Chain Search for the Simple Plant Location Problem improves upper bounds on twelve Körkel-Ghosh instances
Authors	Daniel Karapetyan, Boris Goldengorin
Abstract	We address a family of hard benchmark instances for the Simple Plant Location Problem (also known as the Uncapacitated Facility Location Problem). The recent attempt by Fischetti et al. to tackle the K"orkel-Ghosh instances resulted in seven new optimal solutions and 22 improved upper bounds. We use automated generation of heuristics to obtain a new algorithm for the Simple Plant Location Problem. In our experiments, our new algorithm matched all the previous best known and optimal solutions, and further improved 12 upper bounds, all within shorter time budgets compared to the previous efforts. Our algorithm design process is split into two phases: (i) development of algorithmic components such as local search procedures and mutation operators, and (ii) composition of a metaheuristic from the available components. Phase (i) requires human expertise and often can be completed by implementing several simple domain-specific routines known from the literature. Phase (ii) is entirely automated by employing the Conditional Markov Chain Search (CMCS) framework. In CMCS, a metaheuristic is flexibly defined by a set of parameters, called configuration. Then the process of composition of a metaheuristic from the algorithmic components is reduced to an optimisation problem seeking the best performing CMCS configuration. We discuss the problem of comparing configurations, and propose a new efficient technique to select the best performing configuration from a large set. To employ this method, we restrict the original CMCS to a simple deterministic case that leaves us with a finite and manageable number of meaningful configurations.
Tasks
Published	2017-11-16
URL	http://arxiv.org/abs/1711.06347v1
PDF	http://arxiv.org/pdf/1711.06347v1.pdf
PWC	https://paperswithcode.com/paper/conditional-markov-chain-search-for-the
Repo
Framework

Objective Classes for Micro-Facial Expression Recognition


Title	Objective Classes for Micro-Facial Expression Recognition
Authors	Adrian K. Davison, Walied Merghani, Moi Hoon Yap
Abstract	Micro-expressions are brief spontaneous facial expressions that appear on a face when a person conceals an emotion, making them different to normal facial expressions in subtlety and duration. Currently, emotion classes within the CASME II dataset are based on Action Units and self-reports, creating conflicts during machine learning training. We will show that classifying expressions using Action Units, instead of predicted emotion, removes the potential bias of human reporting. The proposed classes are tested using LBP-TOP, HOOF and HOG 3D feature descriptors. The experiments are evaluated on two benchmark FACS coded datasets: CASME II and SAMM. The best result achieves 86.35% accuracy when classifying the proposed 5 classes on CASME II using HOG 3D, outperforming the result of the state-of-the-art 5-class emotional-based classification in CASME II. Results indicate that classification based on Action Units provides an objective method to improve micro-expression recognition.
Tasks	Facial Expression Recognition
Published	2017-08-24
URL	http://arxiv.org/abs/1708.07549v2
PDF	http://arxiv.org/pdf/1708.07549v2.pdf
PWC	https://paperswithcode.com/paper/objective-classes-for-micro-facial-expression
Repo
Framework

Improving Dermoscopic Image Segmentation with Enhanced Convolutional-Deconvolutional Networks


Title	Improving Dermoscopic Image Segmentation with Enhanced Convolutional-Deconvolutional Networks
Authors	Yading Yuan, Yeh-Chi Lo
Abstract	Automatic skin lesion segmentation on dermoscopic images is an essential step in computer-aided diagnosis of melanoma. However, this task is challenging due to significant variations of lesion appearances across different patients. This challenge is further exacerbated when dealing with a large amount of image data. In this paper, we extended our previous work by developing a deeper network architecture with smaller kernels to enhance its discriminant capacity. In addition, we explicitly included color information from multiple color spaces to facilitate network training and thus to further improve the segmentation performance. We extensively evaluated our method on the ISBI 2017 skin lesion segmentation challenge. By training with the 2000 challenge training images, our method achieved an average Jaccard Index (JA) of 0.765 on the 600 challenge testing images, which ranked itself in the first place in the challenge
Tasks	Lesion Segmentation, Semantic Segmentation
Published	2017-09-28
URL	http://arxiv.org/abs/1709.09780v1
PDF	http://arxiv.org/pdf/1709.09780v1.pdf
PWC	https://paperswithcode.com/paper/improving-dermoscopic-image-segmentation-with
Repo
Framework

Multi-shot ASP solving with clingo


Title	Multi-shot ASP solving with clingo
Authors	Martin Gebser, Roland Kaminski, Benjamin Kaufmann, Torsten Schaub
Abstract	We introduce a new flexible paradigm of grounding and solving in Answer Set Programming (ASP), which we refer to as multi-shot ASP solving, and present its implementation in the ASP system clingo. Multi-shot ASP solving features grounding and solving processes that deal with continuously changing logic programs. In doing so, they remain operative and accommodate changes in a seamless way. For instance, such processes allow for advanced forms of search, as in optimization or theory solving, or interaction with an environment, as in robotics or query-answering. Common to them is that the problem specification evolves during the reasoning process, either because data or constraints are added, deleted, or replaced. This evolutionary aspect adds another dimension to ASP since it brings about state changing operations. We address this issue by providing an operational semantics that characterizes grounding and solving processes in multi-shot ASP solving. This characterization provides a semantic account of grounder and solver states along with the operations manipulating them. The operative nature of multi-shot solving avoids redundancies in relaunching grounder and solver programs and benefits from the solver’s learning capacities. clingo accomplishes this by complementing ASP’s declarative input language with control capacities. On the declarative side, a new directive allows for structuring logic programs into named and parameterizable subprograms. The grounding and integration of these subprograms into the solving process is completely modular and fully controllable from the procedural side. To this end, clingo offers a new application programming interface that is conveniently accessible via scripting languages.
Tasks
Published	2017-05-27
URL	http://arxiv.org/abs/1705.09811v2
PDF	http://arxiv.org/pdf/1705.09811v2.pdf
PWC	https://paperswithcode.com/paper/multi-shot-asp-solving-with-clingo
Repo
Framework

Convergence Rates of Latent Topic Models Under Relaxed Identifiability Conditions


Title	Convergence Rates of Latent Topic Models Under Relaxed Identifiability Conditions
Authors	Yining Wang
Abstract	In this paper we study the frequentist convergence rate for the Latent Dirichlet Allocation (Blei et al., 2003) topic models. We show that the maximum likelihood estimator converges to one of the finitely many equivalent parameters in Wasserstein’s distance metric at a rate of $n^{-1/4}$ without assuming separability or non-degeneracy of the underlying topics and/or the existence of more than three words per document, thus generalizing the previous works of Anandkumar et al. (2012, 2014) from an information-theoretical perspective. We also show that the $n^{-1/4}$ convergence rate is optimal in the worst case.
Tasks	Topic Models
Published	2017-10-30
URL	http://arxiv.org/abs/1710.11070v2
PDF	http://arxiv.org/pdf/1710.11070v2.pdf
PWC	https://paperswithcode.com/paper/convergence-rates-of-latent-topic-models
Repo
Framework

Weak Adaptive Submodularity and Group-Based Active Diagnosis with Applications to State Estimation with Persistent Sensor Faults


Title	Weak Adaptive Submodularity and Group-Based Active Diagnosis with Applications to State Estimation with Persistent Sensor Faults
Authors	Sze Zheng Yong, Lingyun Gao, Necmiye Ozay
Abstract	In this paper, we consider adaptive decision-making problems for stochastic state estimation with partial observations. First, we introduce the concept of weak adaptive submodularity, a generalization of adaptive submodularity, which has found great success in solving challenging adaptive state estimation problems. Then, for the problem of active diagnosis, i.e., discrete state estimation via active sensing, we show that an adaptive greedy policy has a near-optimal performance guarantee when the reward function possesses this property. We further show that the reward function for group-based active diagnosis, which arises in applications such as medical diagnosis and state estimation with persistent sensor faults, is also weakly adaptive submodular. Finally, in experiments of state estimation for an aircraft electrical system with persistent sensor faults, we observe that an adaptive greedy policy performs equally well as an exhaustive search.
Tasks	Decision Making, Medical Diagnosis
Published	2017-01-24
URL	http://arxiv.org/abs/1701.06731v2
PDF	http://arxiv.org/pdf/1701.06731v2.pdf
PWC	https://paperswithcode.com/paper/weak-adaptive-submodularity-and-group-based
Repo
Framework