Spatial understanding is a fundamental problem with wide-reaching real-world applications. The representation of spatial knowledge is often modeled with spatial templates, i.e., regions of acceptability of two objects under an explicit spatial relationship (e.g., "on," "below," etc.). In contrast with prior work that restricts spatial templates to explicit spatial prepositions (e.g., "glass on table"), here we extend this concept to implicit spatial language, i.e., those relationships (generally actions) for which the spatial arrangement of the objects is only implicitly implied (e.g., "man riding horse"). In contrast with explicit relationships, predicting spatial arrangements from implicit spatial language requires significant common sense spatial understanding. Here, we introduce the task of predicting spatial templates for two objects under a relationship, which can be seen as a spatial question-answering task with a (2D) continuous output ("where is the man w.r.t. a horse when the man is walking the horse?"). We present two simple neural-based models that leverage annotated images and structured text to learn this task. The good performance of these models reveals that spatial locations are to a large extent predictable from implicit spatial language. Crucially, the models attain similar performance in a challenging generalized setting, where the object-relation-object combinations (e.g., "man walking dog") have never been seen before. Next, we go one step further by presenting the models with unseen objects (e.g., "dog"). In this scenario, we show that leveraging word embeddings enables the models to output accurate spatial predictions, proving that the models acquire solid common sense spatial knowledge allowing for such generalization.

Spatio-temporal feature encoding is essential for encoding facial expression dynamics in video sequences. At test time, most spatio-temporal encoding methods assume that a temporally segmented sequence is fed to a learned model, which could require the prediction to wait until the full sequence is available to an auxiliary task that performs the temporal segmentation. This causes a delay in predicting the expression. In an interactive setting, such as affective interactive agents, such delay in the prediction could not be tolerated. Therefore, training a model that can accurately predict the facial expression "on-the-fly" (as they are fed to the system) is essential. In this paper, we propose a new spatio-temporal feature learning method, which would allow prediction with partial sequences. As such, the prediction could be performed on-the-fly. The proposed method utilizes an estimated expression intensity to generate dense labels, which are used to regulate the prediction model training with a novel objective function. As results, the learned spatio-temporal features can robustly predict the expression with partial (incomplete) expression sequences, on-the-fly. Experimental results showed that the proposed method achieved higher recognition rates compared to the state-of-the-art methods on both datasets. More importantly, the results verified that the proposed method improved the prediction frames with partial expression sequence inputs.

Propositional satisfiability (SAT) is at the nucleus of state-of-the-art approaches to a variety of computationally hard problems, one of which is cryptanalysis. Moreover, a number of practical applications of SAT can only be tackled efficiently by identifying and exploiting a subset of formula's variables called backdoor set (or simply backdoors). This paper proposes a new class of backdoor sets for SAT used in the context of cryptographic attacks, namely guess-and-determine attacks. The idea is to identify the best set of backdoor variables subject to a statistically estimated hardness of the guess-and-determine attack using a SAT solver. Experimental results on weakened variants of the renowned encryption algorithms exhibit advantage of the proposed approach compared to the state of the art in terms of the estimated hardness of the resulting guess-and-determine attacks.

Methods for explaining the sources of inconsistency of overconstrained systems find an ever-increasing number of applications, ranging from diagnosis and configuration to ontology debugging and axiom pinpointing in description logics. Efficient enumeration of minimal correction subsets (MCSes), defined as sets of constraints whose removal from the system restores feasibility, is a central task in such domains. In this work, we propose a novel approach to speeding up MCS enumeration over conjunctive normal form propositional formulas by caching of so-called premise sets (PSes) seen during the enumeration process. Contrasting to earlier work, we move from caching unsatisfiable cores to caching PSes and propose a more effective way of implementing the cache. The proposed techniques noticeably improves on the performance of state-of-the-art MCS enumeration algorithms in practice.

Multi-valued Decision Diagrams (MDDs) have been extensively studied in the last ten years. Recently, efficient algorithms implementing operators such as reduction, union, intersection, difference, etc., have been designed. They directly deal with the graph structure of the MDD and a time reduction of several orders of magnitude in comparison to other existing algorithms have been observed. These operators have permitted a new look at MDDs, because extremely large MDDs can finally be manipulated as shown by the models used to solve complex application in music generation. However, MDDs become so large (50GB) that minutes are sometimes required to perform some operations. In order to accelerate the manipulation of MDDs, parallel algorithms are required. In this paper, we introduce such algorithms. We carefully design them in order to overcome inherent difficulties of the parallelization of sequential algorithms such as data dependencies, software lock-out, false sharing, or load balancing. As a result, we observe a speed-up , i.e. ratio between parallel and sequential runtimes, growing linearly with the number of cores.

There are well known cases of Quantified Boolean Formulas (QBFs) that have short winning strategies (Skolem/Herbrand functions) but that are hard to solve by nowadays solvers. This paper argues that a solver benefits from generalizing a set of individual wins into a strategy. This idea is realized on top of the competitive RAReQS algorithm by utilizing machine learning, which enables learning shorter strategies. The implemented prototype QFUN has won the first place in the non-CNF track of the most recent QBF competition.

Conflict-driven clause learning (CDCL) is at the core of the success of modern SAT solvers. In terms of propositional proof complexity, CDCL has been shown as strong as general resolution. Improvements to SAT solvers can be realized either by improving existing algorithms, or by exploiting proof systems stronger than CDCL. Recent work proposed an approach for solving SAT by reduction to Horn MaxSAT. The proposed reduction coupled with MaxSAT resolution represents a new proof system, DRMaxSAT, which was shown to enable polynomial time refutations of pigeonhole formulas, in contrast with either CDCL or general resolution. This paper investigates the DRMaxSAT proof system, and shows that DRMaxSAT p-simulates general resolution, that AC0-Frege+PHP p-simulates DRMaxSAT, and that DRMaxSAT can not p-simulate AC0-Frege+PHP or the cutting planes proof system.

Formal verification provides a high degree of confidence in safe system operation, but only if reality matches the verified model. Although a good model will be accurate most of the time, even the best models are incomplete. This is especially true in Cyber-Physical Systems because high-fidelity physical models of systems are expensive to develop and often intractable to verify. Conversely, reinforcement learning-based controllers are lauded for their flexibility in unmodeled environments, but do not provide guarantees of safe operation. This paper presents an approach for provably safe learning that provides the best of both worlds: the exploration and optimization capabilities of learning along with the safety guarantees of formal verification. Our main insight is that formal verification combined with verified runtime monitoring can ensure the safety of a learning agent. Verification results are preserved whenever learning agents limit exploration within the confounds of verified control choices as long as observed reality comports with the model used for off-line verification. When a model violation is detected, the agent abandons efficiency and instead attempts to learn a control strategy that guides the agent to a modeled portion of the state space. We prove that our approach toward incorporating knowledge about safe control into learning systems preserves safety guarantees, and demonstrate that we retain the empirical performance benefits provided by reinforcement learning. We also explore various points in the design space for these justified speculative controllers in a simple model of adaptive cruise control model for autonomous cars.

Given a set of obstacles and two designated points in the plane, the Minimum Constraint Removal problem asks for a minimum number of obstacles that can be removed so that a collision-free path exists between the two designated points. It is a well-studied problem in both robotic motion planning and wireless computing that has been shown to be NP-hard in various settings. In this work, we extend the study of Minimum Constraint Removal. We start by presenting refined NP-hardness reductions for the two cases: (1) when all the obstacles are axes-parallel rectangles, and (2) when all the obstacles are line segments such that no three intersect at the same point. These results improve on existing results in the literature. As a byproduct of our NP-hardness reductions, we prove that, unless the Exponential-Time Hypothesis (ETH) fails, Minimum Constraint Removal cannot be solved in subexponential time 2 o ( n ) , where n is the number of obstacles in the instance. This shows that significant improvement on the brute-force 2 O ( n ) -time algorithm is unlikely. We then present a subexponential-time algorithm for instances of Minimum Constraint Removal in which the number of obstacles that overlap at any point is constant; the algorithm runs in time 2 O (√ N ) , where N is the number of the vertices in the auxiliary graph associated with the instance of the problem. We show that significant improvement on this algorithm is unlikely by showing that, unless ETH fails, Minimum Constraint Removal with bounded overlap number cannot be solved in time 2 o (√ N ) . We describe several exact algorithms and approximation algorithms that leverage heuristics and discuss their performance in an extensive empirical simulation.

Inertial sensors play a pivotal role in indoor localization, which in turn lays the foundation for pervasive personal applications. However, low-cost inertial sensors, as commonly found in smartphones, are plagued by bias and noise, which leads to unbounded growth in error when accelerations are double integrated to obtain displacement. Small errors in state estimation propagate to make odometry virtually unusable in a matter of seconds. We propose to break the cycle of continuous integration, and instead segment inertial data into independent windows. The challenge becomes estimating the latent states of each window, such as velocity and orientation, as these are not directly observable from sensor data. We demonstrate how to formulate this as an optimization problem, and show how deep recurrent neural networks can yield highly accurate trajectories, outperforming state-of-the-art shallow techniques, on a wide range of tests and attachments. In particular, we demonstrate that IONet can generalize to estimate odometry for non-periodic motion, such as a shopping trolley or baby-stroller, an extremely challenging task for existing techniques.