In this paper, we introduce a novel method to interpret recurrent neural networks (RNNs), particularly long short-term memory networks (LSTMs) at the cellular level. We propose a systematic pipeline for interpreting individual hidden state dynamics within the network using response characterization methods. The ranked contribution of individual cells to the network's output is computed by analyzing a set of interpretable metrics of their decoupled step and sinusoidal responses. As a result, our method is able to uniquely identify neurons with insightful dynamics, quantify relationships between dynamical properties and test accuracy through ablation analysis, and interpret the impact of network capacity on a network's dynamical distribution. Finally, we demonstrate generalizability and scalability of our method by evaluating a series of different benchmark sequential datasets.

Abstract--Robust PCA, the problem of PCA in the presence of outliers has been extensively investigated in the last few years. Here we focus on Robust PCA in the outlier model where each column of the data matrix is either an inlier or an outlier. Most of the existing methods for this model assumes either the knowledge of the dimension of the lower dimensional subspace or the fraction of outliers in the system. However in many applications knowledge of these parameters is not available. Motivated by this we propose a parameter free outlier identification method for robust PCA which a) does not require the knowledge of outlier fraction, b) does not require the knowledge of the dimension of the underlying subspace, c) is computationally simple and fast d) can handle structured and unstructured outliers. Further, analytical guarantees are derived for outlier identification and the performance of the algorithm is compared with the existing state of the art methods in both real and synthetic data for various outlier structures. Principal Component Analysis (PCA) [1] is a very widely used technique in data analysis and dimensionality reduction. Singular Value Decomposition (SVD) of the data matrix M [2] is known to be very sensitive to extreme corruptions in the data [3], [4], [5] and hence robustifying the PCA process becomes a necessity. Robust PCA is typically an ill posed problem and it is of significant importance in a wide variety of fields like computer vision, machine learning, survey data analysis and so on. Of the numerous approaches to robust PCA over the years [8], [9], one way to model extreme corruptions in the given data matrix M, is using the following decomposition [10], [11], [12], [3]: M L S, where S encapsulates all the corruptions and is assumed to be sparse and L is low rank.

Sign information is the key to overcoming the inevitable saturation error in compressive sensing systems, which causes information loss and results in bias. For sparse signal recovery from saturation, we propose to use a linear loss to improve the effectiveness from existing methods that utilize hard constraints/hinge loss for sign consistency. Due to the use of linear loss, an analytical solution in the update progress is obtained, and some nonconvex penalties are applicable, e.g., the minimax concave penalty, the l Theoretical analysis reveals that the estimation error can still be bounded. Generally, with linear loss and nonconvex penalties, the recovery performance is significantly improved, and the computational time is largely saved, which is verified by the numerical experiments. When there are saturated measurements, the observation is nonlinear, and the performance of algorithms using linear observations degrades.

Robot manipulation is increasingly poised to interact with humans in co-shared workspaces. Despite increasingly robust manipulation and control algorithms, failure modes continue to exist whenever models do not capture the dynamics of the unstructured environment. To obtain longer-term horizons in robot automation, robots must develop introspection and recovery abilities. We contribute a set of recovery policies to deal with anomalies produced by external disturbances as well as anomaly classification through the use of non-parametric statistics with memoized variational inference with scalable adaptation. A recovery critic stands atop of a tightly-integrated, graph-based online motion-generation and introspection system that resolves a wide range of anomalous situations. Policies, skills, and introspection models are learned incrementally and contextually in a task. Two task-level recovery policies: re-enactment and adaptation resolve accidental and persistent anomalies respectively. The introspection system uses non-parametric priors along with Markov jump linear systems and memoized variational inference with scalable adaptation to learn a model from the data. Extensive real-robot experimentation with various strenuous anomalous conditions is induced and resolved at different phases of a task and in different combinations. The system executes around-the-clock introspection and recovery and even elicited self-recovery when misclassifications occurred.

Vulnerable road users (VRUs, i.e. cyclists and pedestrians) will play an important role in future traffic. To avoid accidents and achieve a highly efficient traffic flow, it is important to detect VRUs and to predict their intentions. In this article a holistic approach for detecting intentions of VRUs by cooperative methods is presented. The intention detection consists of basic movement primitive prediction, e.g. standing, moving, turning, and a forecast of the future trajectory. Vehicles equipped with sensors, data processing systems and communication abilities, referred to as intelligent vehicles, acquire and maintain a local model of their surrounding traffic environment, e.g. crossing cyclists. Heterogeneous, open sets of agents (cooperating and interacting vehicles, infrastructure, e.g. cameras and laser scanners, and VRUs equipped with smart devices and body-worn sensors) exchange information forming a multi-modal sensor system with the goal to reliably and robustly detect VRUs and their intentions under consideration of real time requirements and uncertainties. The resulting model allows to extend the perceptual horizon of the individual agent beyond their own sensory capabilities, enabling a longer forecast horizon. Concealments, implausibilities and inconsistencies are resolved by the collective intelligence of cooperating agents. Novel techniques of signal processing and modelling in combination with analytical and learning based approaches of pattern and activity recognition are used for detection, as well as intention prediction of VRUs. Cooperation, by means of probabilistic sensor and knowledge fusion, takes place on the level of perception and intention recognition. Based on the requirements of the cooperative approach for the communication a new strategy for an ad hoc network is proposed.

In this paper we introduce vSTS, a new dataset for measuring textual similarity of sentences using multimodal information. The dataset is comprised by images along with its respectively textual captions. We describe the dataset both quantitatively and qualitatively, and claim that it is a valid gold standard for measuring automatic multimodal textual similarity systems. We also describe the initial experiments combining the multimodal information.

Sometime in the late 1980s, I participated in a meeting organized by the Honda Foundation in which a Japanese professor--I can't remember his name--made the case that the Japanese had more success integrating robots into society because of their country's indigenous Shinto religion, which remains the official national religion of Japan. Followers of Shinto, unlike Judeo-Christian monotheists and the Greeks before them, do not believe that humans are particularly "special." Instead, there are spirits in everything, rather like the Force in Star Wars. Nature doesn't belong to us, we belong to Nature, and spirits live in everything, including rocks, tools, homes, and even empty spaces. The West, the professor contended, has a problem with the idea of things having spirits and feels that anthropomorphism, the attribution of human-like attributes to things or animals, is childish, primitive, or even bad.

The MRes RCA programme provides early and mid-career art and design researchers with the intellectual, technical and professional tools with which to complete high-quality research projects. The programme is a uniquely interdisciplinary degree, and the first to be taught across all four Schools of the RCA. Over a full-time year it offers training in practice and theory-led research methods for critical studies in art and design. As demonstrated by the graduating students' work, MRes RCA supports students from diverse backgrounds. Students come from previous study both in art and design and in related disciplines such as history, political sciences and psychology, and with experience working in the creative industries, as practising architects, designers and artists.

In the first part of this series on the counterintelligence implications of artificial intelligence (AI), I discussed AI and counterintelligence at a high level and described some features of each that I think are particularly relevant to understanding the intersection between the two fields. That general discussion leads naturally to one particular counterintelligence question related to AI: How do we identify, understand and protect our most valuable AI assets? To do that, it is important to remember that AI systems operate as part of a much larger digital ecosystem. My focus here is on AI assets in general rather than particular applications of AI. Obviously, certain AI systems, such as those used in military, intelligence and critical-infrastructure settings, require special attention from a counterintelligence perspective, but I won't focus on those specifically in this post.

The trio of phones will boast other features, such as facial-recognition technology, but their display size stands out--their average screen area, without accounting for a facial-recognition system that juts into the top of the display, is 23% larger than last year's three new phones and 28% bigger than the two models unveiled in 2016. Note: Screens on iPhone X and newer have facial-recognition systems that cut into the display. At a time when people are buying fewer new phones, bigger size brings two advantages. It helps Apple buoy prices and profit margins because it can sell larger phones at a greater markup than it pays suppliers for the larger screens. And it encourages people to use their phones more, helping momentum of Apple's services business, which includes app-store sales and subscriptions to video services like Netflix and HBO.