Abstract--A novel procedure is presented in this paper, for training a deep convolutional and recurrent neural network, taking into account both the available training data set and some information extracted from similar networks trained with other relevant data sets. This information is included in an extended loss function used for the network training, so that the network can have an improved performance when applied to the other data sets, without forgetting the learned knowledge from the original data set. Facial expression and emotion recognition in-the-wild is the test bed application that is used to demonstrate the improved performance achieved using the proposed approach. In this framework, we provide an experimental study on categorical emotion recognition using datasets from a very recent related emotion recognition challenge. Index Terms --deep neural network training; classification; clustering internal representations; extended loss function; domain adaptation; transfer learning; emotion recognition in-the- wild; . Many real life problems are represented by a variety of data sets which may possess different characteristics. In such cases learning to classify correctly one data set does not generalize well in the other sets.

Music relies heavily on self-reference to build structure and meaning. We explore the Transformer architecture (Vaswani et al., 2017) as a generative model for music, as self-attention has shown compelling results on tasks that require long-term structure such as Wikipedia summary generation (Liu et al, 2018). However, timing information is critical for polyphonic music, and Transformer does not explicitly model absolute or relative timing in its structure. To address this challenge, Shaw et al. (2018) introduced relative position representations to self-attention to improve machine translation. However, the formulation was not scalable to longer sequences. We propose an improved formulation which reduces the memory requirements of the relative position computation from $O(l^2d)$ to $O(ld)$, making it possible to train much longer sequences and achieve faster convergence. In experiments on symbolic music we find that relative self-attention substantially improves sample quality for unconditioned generation and is able to generate sequences of lengths longer than those from the training set. When primed with an initial sequence, the model generates continuations that develop the prime coherently and exhibit long-term structure. Relative self-attention can be instrumental in capturing richer relationships within a musical piece.

We address the problem of unsupervised disentanglement of latent representations learnt via deep generative models. In contrast to current approaches that operate on the evidence lower bound (ELBO), we argue that statistical independence in the latent space of VAEs can be enforced in a principled hierarchical Bayesian manner. To this effect, we augment the standard VAE with an inverse-Wishart (IW) prior on the covariance matrix of the latent code. By tuning the IW parameters, we are able to encourage (or discourage) independence in the learnt latent dimensions. Extensive experimental results on a range of datasets (2DShapes, 3DChairs, 3DFaces and CelebA) show our approach to outperform the $\beta$-VAE and is competitive with the state-of-the-art FactorVAE. Our approach achieves significantly better disentanglement and reconstruction on a new dataset (CorrelatedEllipses) which introduces correlations between the factors of variation.

Ensembles of deep neural networks with diverse architectures significantly improve generalization accuracy. However, training such ensembles requires a large amount of computational resources and time as every network in the ensemble has to be separately trained. In practice, this restricts the number of different deep neural network architectures that can be included within an ensemble. We propose a new approach to address this problem. Our approach captures the structural similarity between members of a neural network ensemble and train it only once. Subsequently, this knowledge is transferred to all members of the ensemble using function-preserving transformations. Then, these ensemble networks converge significantly faster as compared to training from scratch. We show through experiments on CIFAR-10, CIFAR-100, and SVHN data sets that our approach can train large and diverse ensembles of deep neural networks achieving comparable accuracy to existing approaches in a fraction of their training time. In particular, our approach trains an ensemble of $100$ variants of deep neural networks with diverse architectures up to $6 \times$ faster as compared to existing approaches. This improvement in training cost grows linearly with the size of the ensemble.

Classification systems typically act in isolation, meaning they are required to implicitly memorize the characteristics of all candidate classes in order to classify. The cost of this is increased memory usage and poor sample efficiency. We propose a model which instead verifies using reference images during the classification process, reducing the burden of memorization. The model uses iterative nondifferentiable queries in order to classify an image. We demonstrate that such a model is feasible to train and can match baseline accuracy while being more parameter efficient. However, we show that finding the correct balance between image recognition and verification is essential to pushing the model towards desired behavior, suggesting that a pipeline of recognition followed by verification is a more promising approach.

Deep neural networks are gaining increasing popularity for the classic text classification task, due to their strong expressive power and less requirement for feature engineering. Despite such attractiveness, neural text classification models suffer from the lack of training data in many real-world applications. Although many semi-supervised and weakly-supervised text classification models exist, they cannot be easily applied to deep neural models and meanwhile support limited supervision types. In this paper, we propose a weakly-supervised method that addresses the lack of training data in neural text classification. Our method consists of two modules: (1) a pseudo-document generator that leverages seed information to generate pseudo-labeled documents for model pre-training, and (2) a self-training module that bootstraps on real unlabeled data for model refinement. Our method has the flexibility to handle different types of weak supervision and can be easily integrated into existing deep neural models for text classification. We have performed extensive experiments on three real-world datasets from different domains. The results demonstrate that our proposed method achieves inspiring performance without requiring excessive training data and outperforms baseline methods significantly.

Over half a million individuals are diagnosed with head and neck cancer each year worldwide. Radiotherapy is an important curative treatment for this disease, but it requires manually intensive delineation of radiosensitive organs at risk (OARs). This planning process can delay treatment commencement. While auto-segmentation algorithms offer a potentially time-saving solution, the challenges in defining, quantifying and achieving expert performance remain. Adopting a deep learning approach, we demonstrate a 3D U-Net architecture that achieves performance similar to experts in delineating a wide range of head and neck OARs. The model was trained on a dataset of 663 deidentified computed tomography (CT) scans acquired in routine clinical practice and segmented according to consensus OAR definitions. We demonstrate its generalisability through application to an independent test set of 24 CT scans available from The Cancer Imaging Archive collected at multiple international sites previously unseen to the model, each segmented by two independent experts and consisting of 21 OARs commonly segmented in clinical practice. With appropriate validation studies and regulatory approvals, this system could improve the effectiveness of radiotherapy pathways.

Medical applications challenge today's text categorization techniques by demanding both high accuracy and ease-of-interpretation. Although deep learning has provided a leap ahead in accuracy, this leap comes at the sacrifice of interpretability. To address this accuracy-interpretability challenge, we here introduce, for the first time, a text categorization approach that leverages the recently introduced Tsetlin Machine. In all brevity, we represent the terms of a text as propositional variables. From these, we capture categories using simple propositional formulae, such as: if "rash" and "reaction" and "penicillin" then Allergy. The Tsetlin Machine learns these formulae from a labelled text, utilizing conjunctive clauses to represent the particular facets of each category. Indeed, even the absence of terms (negated features) can be used for categorization purposes. Our empirical results are quite conclusive. The Tsetlin Machine either performs on par with or outperforms all of the evaluated methods on both the 20 Newsgroups and IMDb datasets, as well as on a non-public clinical dataset. On average, the Tsetlin Machine delivers the best recall and precision scores across the datasets. The GPU implementation of the Tsetlin Machine is further 8 times faster than the GPU implementation of the neural network. We thus believe that our novel approach can have a significant impact on a wide range of text analysis applications, forming a promising starting point for deeper natural language understanding with the Tsetlin Machine.

We propose a data-efficient Gaussian process-based Bayesian approach to the semi-supervised learning problem on graphs. The proposed model shows extremely competitive performance when compared to the state-of-the-art graph neural networks on semi-supervised learning benchmark experiments, and outperforms the neural networks in active learning experiments where labels are scarce. Furthermore, the model does not require a validation data set for early stopping to control over-fitting. Our model can be viewed as an instance of empirical distribution regression weighted locally by network connectivity. We further motivate the intuitive construction of the model with a Bayesian linear model interpretation where the node features are filtered by an operator related to the graph Laplacian. The method can be easily implemented by adapting off-the-shelf scalable variational inference algorithms for Gaussian processes.

Abstract--In this study, we investigate the limits of the current state of the art AI system for detecting buffer overflows and compare it with current static analysis tools. To do so, we developed a code generator, sbAbI, capable of producing an arbitrarily large number of code samples of controlled complexity. We found that the static analysis engines we examined have good precision, but poor recall on this dataset, except for a sound static analyzer that has good precision and recall. We found that the state of the art AI system, a memory network modeled after Choi et al. [1], can achieve similar performance to the static analysis engines, but requires an exhaustive amount of training data in order to do so. Our work points towards future approaches that may solve these problems; namely, using representations of code that can capture appropriate scope information and using deep learning methods that are able to perform arithmetic operations. Predicting security defects in source code is of significant national security interest. It is ideal to detect security defects during development, before the code is ever run to expose those defects. The current best methods to find security defects before running code are static analysis tools, a variety of which exist and model software in different ways that are all useful for different kinds of flaws. Developers of static analyzers carefully equip them with rules about program behavior, which are used to reason about the safety of the program if it were to run. However, static analyzers are known to be insufficient at finding flaws. The Juliet Test Suite [2]-[4] is a collection of synthetic code containing intentional security defects across hundreds of vulnerabilities in the Common Weakness Enumeration standard, labeled at the line-of-code level. Even state-ofthe art static analyzers perform poorly at finding the defects in Juliet, issuing too many false positives and also too many false negatives [5]-[8].