Dial locks are commonly used to secure a person’s items. Commercially available dial locks often use four or five wheels of letters, allowing a user to select a word as a combination. In order to evaluate the security of these locks, we create a game, with an instance created by the lock designer, and played by a lock owner and a thief. In the game, the lock owner chooses a word as a combination, and the thief creates a brute force strategy to try all possible combinations that yield words until the combination is found. To accomplish the task, the thief will solve a version of the Probabilistic Travelling Salesman Problem (PTSP) by creating an a priori tour through all the words a lock can create. The goal for the game designer, then, is to create a lock configuration that maximizes the expected length of the best possible PTSP tour. This paper describes a Genetic Algorithm (GA) approach to design a near-optimal game, i.e. a lock configuration that makes it as difficult for the thief to crack. An analysis of the output of the GA shows that the locks that the system creates are significantly more secure than both commercial locks, in the context of this game.

Capsule networks have been shown to be powerful models for image classification, thanks to their ability to represent and capture viewpoint variations of an object. However, the high computational complexity of capsule networks that stems from the recurrent dynamic routing poses a major drawback making their use for large-scale image classification challenging. In this work, we propose Star-Caps a capsule-based network that exploits a straight-through attentive routing to address the drawbacks of capsule networks. By utilizing attention modules augmented by differentiable binary routers, the proposed mechanism estimates the routing coefficients between capsules without recurrence, as opposed to prior related work. Subsequently, the routers utilize straight-through estimators to make binary decisions to either connect or disconnect the route between capsules, allowing stable and faster performance. The experiments conducted on several image classification datasets, including MNIST, SmallNorb, CIFAR-10, CIFAR-100, and ImageNet show that Star-Caps outperforms the baseline capsule networks.

We introduce an architecture for image categorization that enables the end-to-end learning of separate visual features for the different classes to distinguish. The proposed model consists of a deep CNN shaped like a tree. The stem of the tree includes a sequence of convolutional layers common to all classes. The stem then splits into multiple branches implementing parallel feature extractors, which are ultimately connected to the final classification layer via learned gated connections. These learned gates determine for each individual class the subset of features to use. Such a scheme naturally encourages the learning of a heterogeneous set of specialized features through the separate branches and it allows each class to use the subset of features that are optimal for its recognition. We show the generality of our proposed method by reshaping several popular CNNs from the literature into our proposed architecture. Our experiments on the CIFAR100, CIFAR10, ImageNet, and Synth datasets show that in each case our resulting model yields a substantial improvement in accuracy over the original CNN. Our empirical analysis also suggests that our scheme acts as a form of beneficial regularization improving generalization performance.