Generative models are now used to create a variety of high-quality digital artifacts. Yet their use in designing physical objects has received far less attention. In this paper, we advocate for the construction toy, LEGO, as a platform for developing generative models of sequential assembly. We develop a generative model based on graph-structured neural networks that can learn from human-built structures and produce visually compelling designs. Our code is released at: https://github.

We introduce a new programming paradigm called oracle-guided decision programming in which a program specifies a Markov Decision Process (MDP) and the language provides a universal policy. We prototype a new programming language, Dodona, that manifests this paradigm using a primitive 'choose' representing nondeterministic choice. The Dodona interpreter returns either a value or a choicepoint that includes a lossless encoding of all information necessary in principle to make an optimal decision. Meta-interpreters query Dodona's (neural) oracle on these choicepoints to get policy and value estimates, which they can use to perform heuristic search on the underlying MDP. We demonstrate Dodona's potential for zero-shot heuristic guidance by meta-learning over hundreds of synthetic tasks that simulate basic operations over lists, trees, Church datastructures, polynomials, first-order terms and higher-order terms.

With the gradual application of infrared night vision vehicle assistance system in automatic driving, the accuracy of the collected infrared images of pedestrians is gradually improved. In this paper, the migration learning method is used to apply YOLOv3 model to realize pedestrian target detection in infrared images. The target detection model YOLOv3 is migrated to the CVC infrared pedestrian data set, and Diou loss is used to replace the loss function of the original YOLO model to test different super parameters to obtain the best migration learning effect. The experimental results show that in the pedestrian detection task of CVC data set, the average accuracy (AP) of Yolov3 model reaches 96.35%, and that of Diou-Yolov3 model is 72.14%, but the latter has a faster convergence rate of loss curve. The effect of migration learning can be obtained by comparing the two models.

By using deep learning approaches, Speech Emotion Recog-nition (SER) on a single domain has achieved many excellentresults. However, cross-domain SER is still a challenging taskdue to the distribution shift between source and target domains.In this work, we propose a Domain Adversarial Neural Net-work (DANN) based approach to mitigate this distribution shiftproblem for cross-lingual SER. Specifically, we add a languageclassifier and gradient reversal layer after the feature extractor toforce the learned representation both language-independent andemotion-meaningful. Our method is unsupervised, i. e., labelson target language are not required, which makes it easier to ap-ply our method to other languages. Experimental results showthe proposed method provides an average absolute improve-ment of 3.91% over the baseline system for arousal and valenceclassification task. Furthermore, we find that batch normaliza-tion is beneficial to the performance gain of DANN. Thereforewe also explore the effect of different ways of data combinationfor batch normalization.

Flooding results in 8 billion dollars of damage annually in the US and causes the most deaths of any weather related event. Due to climate change scientists expect more heavy precipitation events in the future. However, no current datasets exist that contain both hourly precipitation and river flow data. We introduce a novel hourly river flow and precipitation dataset and a second subset of flash flood events with damage estimates and injury counts. Using these datasets we create two challenges (1) general stream flow forecasting and (2) flash flood damage estimation. We have created several publicly available benchmarks and an easy to use package. Additionally, in the future we aim to augment our dataset with snow pack data and soil index moisture data to improve predictions.

Off-the-shelf biomedical embeddings obtained from the recently released various pre-trained language models (such as BERT, XLNET) have demonstrated state-of-the-art results (in terms of accuracy) for the various natural language understanding tasks (NLU) in the biomedical domain. Relation Classification (RC) falls into one of the most critical tasks. In this paper, we explore how to incorporate domain knowledge of the biomedical entities (such as drug, disease, genes), obtained from Knowledge Graph (KG) Embeddings, for predicting Drug-Drug Interaction from textual corpus. We propose a new method, BERTKG-DDI, to combine drug embeddings obtained from its interaction with other biomedical entities along with domain-specific BioBERT embedding-based RC architecture. Experiments conducted on the DDIExtraction 2013 corpus clearly indicate that this strategy improves other baselines architectures by 4.1% macro F1-score.

Convolutional Neural Networks (CNN) has successfully been used to classify diabetic retinopathy (DR) fundus images in recent times. However, deeper representations in CNN only capture higher-level semantics at the expense of losing spatial information. To make predictions very usable for ophthalmologists, we use a post-attention technique called Gradient-weighted Class Activation Mapping (Grad-CAM) on the penultimate layer of deep learning models to produce coarse localisation maps on DR fundus images. This is to help identify discriminative regions in the images, consequently providing enough evidence for ophthalmologists to make a diagnosis and saving lives by early diagnosis. Specifically, this study uses pre-trained weights from four (4) state-of-the-art deep learning models to produce and compare the localisation maps of DR fundus images. The models used include VGG16, ResNet50, InceptionV3, and InceptionResNetV2. We find that InceptionV3 achieves the best performance with a test classification accuracy of 96.07% and localise lesions better and faster than the other models.

Classifiers that are linear in their parameters, and trained by optimizing a convex loss function, have predictable behavior with respect to changes in the training data, initial conditions, and optimization. Such desirable properties are absent in deep neural networks (DNNs), typically trained by non-linear fine-tuning of a pre-trained model. Previous attempts to linearize DNNs have led to interesting theoretical insights, but have not impacted the practice due to the substantial performance gap compared to standard non-linear optimization. We present the first method for linearizing a pre-trained model that achieves comparable performance to non-linear fine-tuning on most of real-world image classification tasks tested, thus enjoying the interpretability of linear models without incurring punishing losses in performance. LQF consists of simple modifications to the architecture, loss function and optimization typically used for classification: Leaky-ReLU instead of ReLU, mean squared loss instead of cross-entropy, and pre-conditioning using Kronecker factorization. None of these changes in isolation is sufficient to approach the performance of non-linear fine-tuning. When used in combination, they allow us to reach comparable performance, and even superior in the low-data regime, while enjoying the simplicity, robustness and interpretability of linear-quadratic optimization.

Current large-scale auto-regressive language models (Radford et al., 2019; Liu et al., 2018; Graves, 2013) display impressive fluency and can generate convincing text. In this work we start by asking the question: Can the generations of these models be reliably distinguished from real text by statistical discriminators? We find experimentally that the answer is affirmative when we have access to the training data for the model, and guardedly affirmative even if we do not. This suggests that the auto-regressive models can be improved by incorporating the (globally normalized) discriminators into the generative process. We give a formalism for this using the Energy-Based Model framework, and show that it indeed improves the results of the generative models, measured both in terms of perplexity and in terms of human evaluation.

We propose a generic reward shaping approach for improving the Similar approaches appear to have worked in literature on container rate of convergence in reinforcement learning (RL), called Self loading [27] and railway scheduling [11] problems, without Improvement Based REwards, or SIBRE. The approach is designed being formally proposed or analysed. One study on bin packing for use in conjunction with any existing RL algorithm, and consists does propose reward shaping explicitly, and is described below. of rewarding improvement over the agent's own past performance. Literature on formal reward shaping: The proposed approach We prove that SIBRE converges in expectation under the same (SIBRE) falls under the category of reward shaping approaches conditions as the original RL algorithm. The reshaped rewards for RL, but with some key novelty points as described help discriminate between policies when the original rewards are below. Prior literature has shown that the optimal policy learnt weakly discriminated or sparse. Experiments on several well-known by RL remains invariant under reward shaping if the modification benchmark environments with different RL algorithms show that can be expressed as a potential function [15].