Search results for "adversarial"
showing 10 items of 30 documents
Crowd-Averse Robust Mean-Field Games: Approximation via State Space Extension
2016
We consider a population of dynamic agents, also referred to as players. The state of each player evolves according to a linear stochastic differential equation driven by a Brownian motion and under the influence of a control and an adversarial disturbance. Every player minimizes a cost functional which involves quadratic terms on state and control plus a cross-coupling mean-field term measuring the congestion resulting from the collective behavior, which motivates the term “crowd-averse.” Motivations for this model are analyzed and discussed in three main contexts: a stock market application, a production engineering example, and a dynamic demand management problem in power systems. For th…
Improving Speaker-Independent Lipreading with Domain-Adversarial Training
2017
We present a Lipreading system, i.e. a speech recognition system using only visual features, which uses domain-adversarial training for speaker independence. Domain-adversarial training is integrated into the optimization of a lipreader based on a stack of feedforward and LSTM (Long Short-Term Memory) recurrent neural networks, yielding an end-to-end trainable system which only requires a very small number of frames of untranscribed target data to substantially improve the recognition accuracy on the target speaker. On pairs of different source and target speakers, we achieve a relative accuracy improvement of around 40% with only 15 to 20 seconds of untranscribed target speech data. On mul…
Thompson Sampling Guided Stochastic Searching on the Line for Non-stationary Adversarial Learning
2015
This paper reports the first known solution to the N-Door puzzle when the environment is both non-stationary and deceptive (adversarial learning). The Multi-Armed-Bandit (MAB) problem is the iconic representation of the exploration versus exploitation dilemma. In brief, a gambler repeatedly selects and play, one out of N possible slot machines or arms and either receives a reward or a penalty. The objective of the gambler is then to locate the most rewarding arm to play, while in the process maximize his winnings. In this paper we investigate a challenging variant of the MAB problem, namely the non-stationary N-Door puzzle. Here, instead of directly observing the reward, the gambler is only…
Robust consensus in social networks and coalitional games
2014
We study an n-player averaging process with dynamics subject to controls and adversarial disturbances. The model arises in two distinct application domains: i) coalitional games with transferable utilities (TU) and ii) opinion propagation. We study conditions under which the average allocations achieve robust consensus to some predefined target set.
Cloning and training collective intelligence with generative adversarial networks
2021
Industry 4.0 and highly automated critical infrastructure can be seen as cyber‐physical‐social systems controlled by the Collective Intelligence. Such systems are essential for the functioning of the society and economy. On one hand, they have flexible infrastructure of heterogeneous systems and assets. On the other hand, they are social systems, which include collaborating humans and artificial decision makers. Such (human plus machine) resources must be pre‐trained to perform their mission with high efficiency. Both human and machine learning approaches must be bridged to enable such training. The importance of these systems requires the anticipation of the potential and previously unknow…
Crowd-Averse Cyber-Physical Systems: The Paradigm of Robust Mean-Field Games
2016
For a networked controlled system, we illustrate the paradigm of robust mean-field games. This is a modeling framework at the interface of differential game theory, mathematical physics, and $H_{\infty}$ - optimal control that tries to capture the mutual influence between a crowd and its individuals. First, we establish a mean-field system for such games including the effects of adversarial disturbances. Second, we identify the optimal response of the individuals for a given population behavior. Third, we provide an analysis of equilibria and their stability.
Clinically-Driven Virtual Patient Cohorts Generation: An Application to Aorta
2021
The combination of machine learning methods together with computational modeling and simulation of the cardiovascular system brings the possibility of obtaining very valuable information about new therapies or clinical devices through in-silico experiments. However, the application of machine learning methods demands access to large cohorts of patients. As an alternative to medical data acquisition and processing, which often requires some degree of manual intervention, the generation of virtual cohorts made of synthetic patients can be automated. However, the generation of a synthetic sample can still be computationally demanding to guarantee that it is clinically meaningful and that it re…
Countering Adversarial Inference Evasion Attacks Towards ML-Based Smart Lock in Cyber-Physical System Context
2021
Machine Learning (ML) has been taking significant evolutionary steps and provided sophisticated means in developing novel and smart, up-to-date applications. However, the development has also brought new types of hazards into the daylight that can have even destructive consequences required to be addressed. Evasion attacks are among the most utilized attacks that can be generated in adversarial settings during the system operation. In assumption, ML environment is benign, but in reality, perpetrators may exploit vulnerabilities to conduct these gradient-free or gradient-based malicious adversarial inference attacks towards cyber-physical systems (CPS), such as smart buildings. Evasion attac…
Cross-Sensor Adversarial Domain Adaptation of Landsat-8 and Proba-V images for Cloud Detection
2021
The number of Earth observation satellites carrying optical sensors with similar characteristics is constantly growing. Despite their similarities and the potential synergies among them, derived satellite products are often developed for each sensor independently. Differences in retrieved radiances lead to significant drops in accuracy, which hampers knowledge and information sharing across sensors. This is particularly harmful for machine learning algorithms, since gathering new ground truth data to train models for each sensor is costly and requires experienced manpower. In this work, we propose a domain adaptation transformation to reduce the statistical differences between images of two…
Enforcing Perceptual Consistency on Generative Adversarial Networks by Using the Normalised Laplacian Pyramid Distance
2019
In recent years there has been a growing interest in image generation through deep learning. While an important part of the evaluation of the generated images usually involves visual inspection, the inclusion of human perception as a factor in the training process is often overlooked. In this paper we propose an alternative perceptual regulariser for image-to-image translation using conditional generative adversarial networks (cGANs). To do so automatically (avoiding visual inspection), we use the Normalised Laplacian Pyramid Distance (NLPD) to measure the perceptual similarity between the generated image and the original image. The NLPD is based on the principle of normalising the value of…