It has been suggested that adversarial examples cause deep learning models to make incorrect predictions with high confidence. In this work, we take the opposite stance: an overly confident model is more likely to be vulnerable to adversarial examples. This work is one of the most proactive approaches taken to date, as we link robustness with non-calibrated model confidence on noisy images, providing a data-augmentation-free path forward. The adversarial examples phenomenon is most easily explained by the trend of increasing non-regularized model capacity, while the diversity and number of samples in common datasets has remained flat. Test accuracy has incorrectly been associated with true generalization performance, ignoring that training and test splits are often extremely similar in terms of the overall representation space. The transferability property of adversarial examples was previously used as evidence against overfitting arguments, a perceived random effect, but overfitting is not always random.

The learnability of different neural architectures can be characterized directly by computable measures of data complexity. In this paper, we reframe the problem of architecture selection as understanding how data determines the most expressive and generalizable architectures suited to that data, beyond inductive bias. After suggesting algebraic topology as a measure for data complexity, we show that the power of a network to express the topological complexity of a dataset in its decision region is a strictly limiting factor in its ability to generalize. We then provide the first empirical characterization of the topological capacity of neural networks. Our empirical analysis shows that at every level of dataset complexity, neural networks exhibit topological phase transitions. This observation allowed us to connect existing theory to empirically driven conjectures on the choice of architectures for fully-connected neural networks.

A major challenge in understanding the generalization of deep learning is to explain why (stochastic) gradient descent can exploit the network architecture to find solutions that have good generalization performance when using high capacity models. We find simple but realistic examples showing that this phenomenon exists even when learning linear classifiers --- between two linear networks with the same capacity, the one with a convolutional layer can generalize better than the other when the data distribution has some underlying spatial structure. We argue that this difference results from a combination of the convolution architecture, data distribution and gradient descent, all of which are necessary to be included in a meaningful analysis. We provide a general analysis of the generalization performance as a function of data distribution and convolutional filter size, given gradient descent as the optimization algorithm, then interpret the results using concrete examples. Experimental results show that our analysis is able to explain what happens in our introduced examples.

Graph embedding is an effective method to represent graph data in a low dimensional space for graph analytics. Most existing embedding algorithms typically focus on preserving the topological structure or minimizing the reconstruction errors of graph data, but they have mostly ignored the data distribution of the latent codes from the graphs, which often results in inferior embedding in real-world graph data. In this paper, we propose a novel adversarial graph embedding framework for graph data. The framework encodes the topological structure and node content in a graph to a compact representation, on which a decoder is trained to reconstruct the graph structure. Furthermore, the latent representation is enforced to match a prior distribution via an adversarial training scheme. To learn a robust embedding, two variants of adversarial approaches, adversarially regularized graph autoencoder (ARGA) and adversarially regularized variational graph autoencoder (ARVGA), are developed. Experimental studies on real-world graphs validate our design and demonstrate that our algorithms outperform baselines by a wide margin in link prediction, graph clustering, and graph visualization tasks.

Learning high quality class representations from few examples is a key problem in metric-learning approaches to few-shot learning. To accomplish this, we introduce a novel architecture where class representations are conditioned for each few-shot trial based on a target image. We also deviate from traditional metric-learning approaches by training a network to perform comparisons between classes rather than relying on a static metric comparison. This allows the network to decide what aspects of each class are important for the comparison at hand. We find that this flexible architecture works well in practice, achieving state-of-the-art performance on the Caltech-UCSD birds fine-grained classification task.

Modern reinforcement learning algorithms reach super-human performance in many board and video games, but they are sample inefficient, i.e. they typically require significantly more playing experience than humans to reach an equal performance level. To improve sample efficiency, an agent may build a model of the environment and use planning methods to update its policy. In this article we introduce VaST (Variational State Tabulation), which maps an environment with a high-dimensional state space (e.g. the space of visual inputs) to an abstract tabular environment. Prioritized sweeping with small backups, a highly efficient planning method, can then be used to update state-action values. We show how VaST can rapidly learn to maximize reward in tasks like 3D navigation and efficiently adapt to sudden changes in rewards or transition probabilities.

An important frontier in the quest for human-like AI is compositional semantics: how do we design systems that understand an infinite number of expressions built from a finite vocabulary? Recent research has attempted to solve this problem by using deep neural networks to learn vector space embeddings of sentences, which then serve as input to supervised learning problems like paraphrase detection and sentiment analysis. Here we focus on 'natural language inference' (NLI) as a critical test of a system's capacity for semantic compositionality. In the NLI task, sentence pairs are assigned one of three categories: entailment, contradiction, or neutral. We present a new set of NLI sentence pairs that cannot be solved using only word-level knowledge and instead require some degree of compositionality. We use state of the art sentence embeddings trained on NLI (InferSent, Conneau et al. (2017)), and find that performance on our new dataset is poor, indicating that the representations learned by this model fail to capture the needed compositionality. We analyze some of the decision rules learned by InferSent and find that they are largely driven by simple heuristics at the word level that are ecologically valid in the SNLI dataset on which InferSent is trained. Further, we find that augmenting the training dataset with our new dataset improves performance on a held-out test set without loss of performance on the SNLI test set. This highlights the importance of structured datasets in better understanding, as well as improving the performance of, AI systems.

Existing deep multitask learning (MTL) approaches align layers shared between tasks in a parallel ordering. Such an organization significantly constricts the types of shared structure that can be learned. The necessity of parallel ordering for deep MTL is first tested by comparing it with permuted ordering of shared layers. The results indicate that a flexible ordering can enable more effective sharing, thus motivating the development of a soft ordering approach, which learns how shared layers are applied in different ways for different tasks. Deep MTL with soft ordering outperforms parallel ordering methods across a series of domains. These results suggest that the power of deep MTL comes from learning highly general building blocks that can be assembled to meet the demands of each task.

We propose an efficient sparse coding (SC) framework for obtaining sparse representation of data. The proposed framework is very general and applies to both the single dictionary setting, where each data point is represented as a sparse combination of the columns of one dictionary matrix, as well as the multiple dictionary setting as given in morphological component analysis (MCA), where the goal is to separate the data into additive parts such that each part has distinct sparse representation within an appropriately chosen corresponding dictionary. Both tasks have been cast as $\ell_1$-regularized optimization problems of minimizing quadratic reconstruction error. In an effort to accelerate traditional acquisition of sparse codes, we propose a deep learning architecture that constitutes a trainable time-unfolded version of the Split Augmented Lagrangian Shrinkage Algorithm (SALSA), a special case of the alternating direction method of multipliers (ADMM). We empirically validate both variants of the algorithm on image vision tasks and demonstrate that at inference our networks achieve improvements in terms of the running time and the quality of estimated sparse codes on both classic SC and MCA problems over more common baselines. We finally demonstrate the visual advantage of our technique on the task of source separation.

Abstract-- Autonomous driving requires 3D perception of vehicles and other objects in the in environment. Much of the current methods support 2D vehicle detection. This paper proposes a flexible pipeline to adopt any 2D detection network and fuse it with a 3D point cloud to generate 3D information with minimum changes of the 2D detection networks. To identify the 3D box, an effective model fitting algorithm is developed based on generalised car models and score maps. A two-stage convolutional neural network (CNN) is proposed to refine the detected 3D box. This pipeline is tested on the KITTI dataset using two different 2D detection networks. The 3D detection results based on these two networks are similar, demonstrating the flexibility of the proposed pipeline. The results rank second among the 3D detection algorithms, indicating its competencies in 3D detection. I. INTRODUCTION Vision-based car detection has been well developed and widely implemented using deep learning technologies. The KITTI [1] benchmark site reports that the state of the art algorithms are able to achieve 90% average precision (AP). However, for autonomous vehicles, car detection in 2D images is not sufficient to provide enough information for the vehicle to perform planning and decision making due to the lack of depth data.