Despite their success in massive engineering applications, deep neural networks are vulnerable to various perturbations due to their black-box nature. Recent study has shown that a deep neural network can misclassify the data even if the input data is perturbed by an imperceptible amount. In this paper, we address the robustness issue of neural networks by a novel close-loop control method from the perspective of dynamic systems. Instead of modifying the parameters in a fixed neural network architecture, a close-loop control process is added to generate control signals adaptively for the perturbed or corrupted data. We connect the robustness of neural networks with optimal control using the geometrical information of underlying data to design the control objective. The detailed analysis shows how the embedding manifolds of state trajectory affect error estimation of the proposed method. Our approach can simultaneously maintain the performance on clean data and improve the robustness against many types of data perturbations. It can also further improve the performance of robustly trained neural networks against different perturbations. To the best of our knowledge, this is the first work that improves the robustness of neural networks with close-loop control.

An emerging amount of intelligent applications have been developed with the surge of Machine Learning (ML). Deep Neural Networks (DNNs) have demonstrated unprecedented performance across various fields such as medical diagnosis and autonomous driving. While DNNs are widely employed in security-sensitive fields, they are identified to be vulnerable to Neural Trojan (NT) attacks that are controlled and activated by the stealthy trigger. We call this vulnerable model adversarial artificial intelligence (AI). In this paper, we target to design a robust Trojan detection scheme that inspects whether a pre-trained AI model has been Trojaned before its deployment. Prior works are oblivious of the intrinsic property of trigger distribution and try to reconstruct the trigger pattern using simple heuristics, i.e., stimulating the given model to incorrect outputs. As a result, their detection time and effectiveness are limited. We leverage the observation that the pixel trigger typically features spatial dependency and propose TAD, the first trigger approximation based Trojan detection framework that enables fast and scalable search of the trigger in the input space. Furthermore, TAD can also detect Trojans embedded in the feature space where certain filter transformations are used to activate the Trojan. We perform extensive experiments to investigate the performance of the TAD across various datasets and ML models. Empirical results show that TAD achieves a ROC-AUC score of 0:91 on the public TrojAI dataset 1 and the average detection time per model is 7:1 minutes.

There is a clear need to involve patients in medical decisions. However, cognitive psychological research has highlighted the cognitive limitations of humans with respect to 1. Probabilistic assessment of the patient state and of potential outcomes of various decisions, 2. Elicitation of the patient utility function, and 3. Integration of the probabilistic knowledge and of patient preferences to determine the optimal strategy. Therefore, without adequate computational support, current shared decision models have severe ethical deficiencies. An informed consent model unfairly transfers the responsibility to a patient who does not have the necessary knowledge, nor the integration capability. A paternalistic model endows with exaggerated power a physician who might not be aware of the patient preferences, is prone to multiple cognitive biases, and whose computational integration capability is bounded. Recent progress in Artificial Intelligence suggests adding a third agent: a computer, in all deliberative medical decisions: Non emergency medical decisions in which more than one alternative exists, the patient preferences can be elicited, the therapeutic alternatives might be influenced by these preferences, medical knowledge exists regarding the likelihood of the decision outcomes, and there is sufficient decision time. Ethical physicians should exploit computational decision support technologies, neither making the decisions solely on their own, nor shirking their duty and shifting the responsibility to patients in the name of informed consent. The resulting three way (patient, care provider, computer) human machine model that we suggest emphasizes the patient preferences, the physician knowledge, and the computational integration of both aspects, does not diminish the physician role, but rather brings out the best in human and machine.

Recent advancements in machine learning and signal processing domains have resulted in an extensive surge of interest in Deep Neural Networks (DNNs) due to their unprecedented performance and high accuracy for different and challenging problems of significant engineering importance. However, when such deep learning architectures are utilized for making critical decisions such as the ones that involve human lives (e.g., in control systems and medical applications), it is of paramount importance to understand, trust, and in one word "explain" the argument behind deep models' decisions. In many applications, artificial neural networks (including DNNs) are considered as black-box systems, which do not provide sufficient clue on their internal processing actions. Although some recent efforts have been initiated to explain the behaviors and decisions of deep networks, explainable artificial intelligence (XAI) domain, which aims at reasoning about the behavior and decisions of DNNs, is still in its infancy. The aim of this paper is to provide a comprehensive overview on Understanding, Visualization, and Explanation of the internal and overall behavior of DNNs.

Metasurfaces have provided a novel and promising platform for the realization of compact and large-scale optical devices. The conventional metasurface design approach assumes periodic boundary conditions for each element, which is inaccurate in most cases since the near-field coupling effects between elements will change when surrounded by non-identical structures. In this paper, we propose a deep learning approach to predict the actual electromagnetic (EM) responses of each target meta-atom placed in a large array with near-field coupling effects taken into account. The predicting neural network takes the physical specifications of the target meta-atom and its neighbors as input, and calculates its phase and amplitude in milliseconds. This approach can be applied to explain metasurfaces' performance deterioration caused by mutual coupling and further used to optimize their efficiencies once combined with optimization algorithms. To demonstrate the efficacy of this methodology, we obtain large improvements in efficiency for a beam deflector and a metalens over the conventional design approach. Moreover, we show the correlations between a metasurface's performance and its design errors caused by mutual coupling are not bound to certain specifications (materials, shapes, etc.). As such, we envision that this approach can be readily applied to explore the mutual coupling effects and improve the performance of various metasurface designs.

Multi-scale biomedical knowledge networks are expanding with emerging experimental technologies. Link prediction is increasingly used especially in bipartite biomedical networks. We propose a Graph Neural Networks (GNN) method, namely Graph Pair based Link Prediction model (GPLP), for predicting biomedical network links simply based on their topological interaction information. In GPLP, 1-hop subgraphs extracted from known network interaction matrix is learnt to predict missing links. To evaluate our method, three heterogeneous biomedical networks were used, i.e. Drug-Target Interaction network (DTI), Compound-Protein Interaction network (CPI) from NIH Tox21, and Compound-Virus Inhibition network (CVI). In 5-fold cross validation, our proposed GPLP method significantly outperforms over the state-of-the-art baselines. Besides, robustness is tested with different network incompleteness. Our method has the potential applications in other biomedical networks.

Super-resolution is a fundamental problem in computer vision which aims to overcome the spatial limitation of camera sensors. While significant progress has been made in single image super-resolution, most algorithms only perform well on synthetic data, which limits their applications in real scenarios. In this paper, we study the problem of real-scene single image super-resolution to bridge the gap between synthetic data and real captured images. We focus on two issues of existing super-resolution algorithms: lack of realistic training data and insufficient utilization of visual information obtained from cameras. To address the first issue, we propose a method to generate more realistic training data by mimicking the imaging process of digital cameras. For the second issue, we develop a two-branch convolutional neural network to exploit the radiance information originally-recorded in raw images. In addition, we propose a dense channel-attention block for better image restoration as well as a learning-based guided filter network for effective color correction. Our model is able to generalize to different cameras without deliberately training on images from specific camera types. Extensive experiments demonstrate that the proposed algorithm can recover fine details and clear structures, and achieve high-quality results for single image super-resolution in real scenes.

Nowadays, sensors play a major role in several contexts like science, industry and daily life which benefit of their use. However, the retrieved information must be reliable. Anomalies in the behavior of sensors can give rise to critical consequences such as ruining a scientific project or jeopardizing the quality of the production in industrial production lines. One of the more subtle kind of anomalies are uncalibrations. An uncalibration is said to take place when the sensor is not adjusted or standardized by calibration according to a ground truth value. In this work, an online machine-learning based uncalibration detector for temperature, humidity and pressure sensors was developed. This solution integrates an Artificial Neural Network as main component which learns from the behavior of the sensors under calibrated conditions. Then, after trained and deployed, it detects uncalibrations once they take place. The obtained results show that the proposed solution is able to detect uncalibrations for deviation values of 0.25 degrees, 1% RH and 1.5 Pa, respectively. This solution can be adapted to different contexts by means of transfer learning, whose application allows for the addition of new sensors, the deployment into new environments and the retraining of the model with minimum amounts of data.

Machine Learning (ML) is now used in a range of systems with results that are reported to exceed, under certain conditions, human performance. Many of these systems, in domains such as healthcare, automotive and manufacturing, exhibit high degrees of autonomy and are safety critical. Establishing justified confidence in ML forms a core part of the safety case for these systems. In this document we introduce a methodology for the Assurance of Machine Learning for use in Autonomous Systems (AMLAS). AMLAS comprises a set of safety case patterns and a process for (1) systematically integrating safety assurance into the development of ML components and (2) for generating the evidence base for explicitly justifying the acceptable safety of these components when integrated into autonomous system applications. The material in this document is provided as guidance only. No responsibility for loss occasioned to any person acting or refraining from action as a result of this material or any comments made can be accepted by the authors or The University of York.

As large-scale graphs become increasingly more prevalent, it poses significant computational challenges to process, extract and analyze large graph data. Graph coarsening is one popular technique to reduce the size of a graph while maintaining essential properties. Despite rich graph coarsening literature, there is only limited exploration of data-driven methods in the field. In this work, we leverage the recent progress of deep learning on graphs for graph coarsening. We first propose a framework for measuring the quality of coarsening algorithm and show that depending on the goal, we need to carefully choose the Laplace operator on the coarse graph and associated projection/lift operators. Motivated by the observation that the current choice of edge weight for the coarse graph may be sub-optimal, we parametrize the weight assignment map with graph neural networks and train it to improve the coarsening quality in an unsupervised way. Through extensive experiments on both synthetic and real networks, we demonstrate that our method significantly improves common graph coarsening methods under various metrics, reduction ratios, graph sizes, and graph types. It generalizes to graphs of larger size ($25\times$ of training graphs), is adaptive to different losses (differentiable and non-differentiable), and scales to much larger graphs than previous work.