Rigid origami has shown potential in large diversity of practical applications. However, current rigid origami crease pattern design mostly relies on known tessellations. This strongly limits the diversity and novelty of patterns that can be created. In this work, we build upon the recently developed principle of three units method to formulate rigid origami design as a discrete optimization problem, the rigid origami game. Our implementation allows for a simple definition of diverse objectives and thereby expands the potential of rigid origami further to optimized, application-specific crease patterns. We showcase the flexibility of our formulation through use of a diverse set of search methods in several illustrative case studies. We are not only able to construct various patterns that approximate given target shapes, but to also specify abstract, function-based rewards which result in novel, foldable and functional designs for everyday objects.

In this work, we introduce a novel pipeline that enables the generation of multiple-choice questions and exercises from as little as a topic keyword. Hence, providing users the possibility to start with a study objective in mind and then automatically generate personalized learning material. The main contributions of this project are a scraper that can extract relevant information from websites, a novel distractor generation method that can make use of context and a technique to automatically combine text and questions into interactive exercises. Our novel distractor generation method was tested in a human survey which showed that the distractor generation quality is comparable to hand crafted distractors. The pipeline is built into a web application that lets users refine the results for each step, openly accessible at https://adaptive-teaching.com.

Brain decoding, understood as the process of mapping brain activities to the stimuli that generated them, has been an active research area in the last years. In the case of language stimuli, recent studies have shown that it is possible to decode fMRI scans into an embedding of the word a subject is reading. However, such word embeddings are designed for natural language processing tasks rather than for brain decoding. Therefore, they limit the model’s ability to recover the precise stimulus. In this work, we propose to directly classify an fMRI scan, mapping it to the corresponding word within a fixed vocabulary. Unlike existing work, we evaluate on scans from previously unseen subjects. We argue that this is a more realistic setup and we present a model that can decode fMRI data from unseen subjects with 2.62% Top-1 and 9.76% Top5 accuracy in this challenging task. Moreover our model can be fine-tuned on data from the test subject to achieve 4.22% Top-1 and 12.87% Top-5 accuracy, significantly outperforming all the considered competitive baselines.

Many deep learning methods have been proposed recently to learn algorithms for combinatorial problems. However, most approaches focus on either supervised/imitation learning (the target algorithm is known) or single agent reinforcement learning (the input distribution is fixed). In some cases, however, the input distribution scales combinatorially as well and cannot easily be fully represented in a concise data set. In this paper, we propose a self-play approach to learn a distributed directory protocol to coordinate concurrent requests to a shared mobile resource among a network of nodes. The self-play is between two agents – a request agent, which finds worst case request inputs – and a route agent, which finds an algorithm that works well on the proposed worst case inputs (and consequently, works well on most queries). We show that our self-play approach is successful in learning an algorithm that works well across diverse request sequences. The empirical performance of the learnt algorithms is within the best known theoretical bounds, and sometimes significantly better than the best known upper bounds.

In this paper we study the problem of monaural music source separation, where a piece of music is to be separated into its main constituent sources. We propose a simple yet effective deep neural network architecture based on a ResNet autoencoder. We investigate several data augmentation and post-processing methods to improve the separation results and outperform various state of the art monaural source separation methods on the DSD100 and MUSDB18 datasets. Our results suggest that in order to further push the state of the art in monaural music source separation we need more data, better data augmentation methods, as well as more effective post-processing methods; and not necessarily ever more complex neural network architectures.

The ability to use a 2D map to navigate a complex 3D environment is quite remarkable, and even difficult for many humans. Localization and navigation is also an important problem in domains such as robotics, and has recently become a focus of the deep reinforcement learning community. In this paper we teach a reinforcement learning agent to read a map in order to find the shortest way out of a random maze it has never seen before. Our system combines several state-of-the-art methods such as A3C and incorporates novel elements such as a recurrent localization cell. Our agent learns to localize itself based on 3D first person images and an approximate orientation angle. The agent generalizes well to bigger mazes, showing that it learned useful localization and navigation capabilities.

Learning in sparse reward settings remains a challenge in Reinforcement Learning, which is often addressed by using intrinsic rewards. One promising strategy is inspired by human curiosity, requiring the agent to learn to predict the future. In this paper a curiosity-driven agent is extended to use these predictions directly for training. To achieve this, the agent predicts the value function of the next state at any point in time. Subsequently, the consistency of this prediction with the current value function is measured, which is then used as a regularization term in the loss function of the algorithm. Experiments were made on grid-world environments as well as on a 3D navigation task, both with sparse rewards. In the first case the extended agent is able to learn significantly faster than the baselines.

We present a generally applicable tree-like sparse multilayer architecture that has a balanced connection from all input neurons to all output neurons. If the ratio between input and output neurons is fixed, the parameters required by our architecture scale with O(n1.5) as compared to O(n2) in a fully connected layer, where n is the number of input neurons. Our sparse 2-layer architecture performs similar and/or superior when compared to its fully connected 1-layer and 2-layer counter parts on the IMDB review sentiment classification task, the Reuters news categorization task and the CIFAR-10 image classification task.