Amir Gholami, Kurt Keutzer, George Biros
2019-02-27
Residual neural networks can be viewed as the forward Euler discretization of an Ordinary Differential Equation (ODE) with a unit time step. This has recently motivated researchers to explore other discretization approaches and train ODE based networks. However, an important challenge of neural ODEs is their prohibitive memory cost during gradient backpropogation. Recently a method proposed in [8], claimed that this memory overhead can be reduced from O(LN_t), where N_t is the number of time steps, down to O(L) by solving forward ODE backwards in time, where L is the depth of the network. However, we will show that this approach may lead to several problems: (i) it may be numerically unstable for ReLU/non-ReLU activations and general convolution operators, and (ii) the proposed optimize-then-discretize approach may lead to divergent training due to inconsistent gradients for small time step sizes. We discuss the underlying problems, and to address them we propose ANODE, an Adjoint based Neural ODE framework which avoids the numerical instability related problems noted above, and provides unconditionally accurate gradients. ANODE has a memory footprint of O(L) + O(N_t), with the same computational cost as reversing ODE solve. We furthermore, discuss a memory efficient algorithm which can further reduce this footprint with a trade-off of additional computational cost. We show results on Cifar-10/100 datasets using ResNet and SqueezeNext neural networks.
Weili Nie, Ankit Patel
2018-06-24
Generative adversarial networks (GANs) are notoriously difficult to train and the reasons underlying their (non-)convergence behaviors are still not completely understood. By first considering a simple yet representative GAN example, we mathematically analyze its local convergence behavior in a non-asymptotic way. Furthermore, the analysis is extended to general GANs under certain assumptions. We find that in order to ensure a good convergence rate, two factors of the Jacobian in the GAN training dynamics should be simultaneously avoided, which are (i) the Phase Factor, i.e., the Jacobian has complex eigenvalues with a large imaginary-to-real ratio, and (ii) the Conditioning Factor, i.e., the Jacobian is ill-conditioned. Previous methods of regularizing the Jacobian can only alleviate one of these two factors, while making the other more severe. Thus we propose a new JAcobian REgularization (JARE) for GANs, which simultaneously addresses both factors by construction. Finally, we conduct experiments that confirm our theoretical analysis and demonstrate the advantages of JARE over previous methods in stabilizing GANs.
Antônio H. Ribeiro, Koen Tiels, Luis A. Aguirre, Thomas B. Schön
2019-06-20
Training recurrent neural networks (RNNs) that possess long-term memory is challenging. We provide insight into the trade-off between the smoothness of the cost function and the memory retention capabilities of the network. We express both aspects in terms of the Lipschitz constant of the dynamics modeled by the network. This allows us to make a distinction between three types of regions in the parameter space. In the first region, the network experiences problems in retaining long-term information, while at the same time the cost function is smooth and easy for gradient descent to navigate in. In the second region, the amount of stored information increases with time and the cost function is intricate and full of local minima. The third region is in between the two other regions and here the RNN is able to retain long-term information. Based on these theoretical findings we present the hypothesis that good parameter choices for the RNN are located in between the well-behaved and the ill-behaved cost function regions. The concepts presented in the paper are illustrated by artificially generated and real examples.
Mital Kinderkhedia
2019-06-07
Deep Neural Networks have shown tremendous success in the area of object recognition, image classification and natural language processing. However, designing optimal Neural Network architectures that can learn and output arbitrary graphs is an ongoing research problem. The objective of this survey is to summarize and discuss the latest advances in methods to Learn Representations of Graph Data. We start by identifying commonly used types of graph data and review basics of graph theory. This is followed by a discussion of the relationships between graph kernel methods and neural networks. Next we identify the major approaches used for learning representations of graph data namely: Kernel approaches, Convolutional approaches, Graph neural networks approaches, Graph embedding approaches and Probabilistic approaches. A variety of methods under each of the approaches are discussed and the survey is concluded with a brief discussion of the future of learning representation of graph data.
Guandao Yang, Xun Huang, Zekun Hao, Ming-Yu Liu, Serge Belongie, Bharath Hariharan
2019-06-28
As 3D point clouds become the representation of choice for multiple vision and graphics applications, the ability to synthesize or reconstruct high-resolution, high-fidelity point clouds becomes crucial. Despite the recent success of deep learning models in discriminative tasks of point clouds, generating point clouds remains challenging. This paper proposes a principled probabilistic framework to generate 3D point clouds by modeling them as a distribution of distributions. Specifically, we learn a two-level hierarchy of distributions where the first level is the distribution of shapes and the second level is the distribution of points given a shape. This formulation allows us to both sample shapes and sample an arbitrary number of points from a shape. Our generative model, named PointFlow, learns each level of the distribution with a continuous normalizing flow. The invertibility of normalizing flows enables the computation of the likelihood during training and allows us to train our model in the variational inference framework. Empirically, we demonstrate that PointFlow achieves state-of-the-art performance in point cloud generation. We additionally show that our model can faithfully reconstruct point clouds and learn useful representations in an unsupervised manner. The code will be available at https://github.com/stevenygd/PointFlow.
Simon Carbonnelle, Christophe De Vleeschouwer
2018-06-05
Our work presents extensive empirical evidence that layer rotation, i.e. the evolution across training of the cosine distance between each layer's weight vector and its initialization, constitutes an impressively consistent indicator of generalization performance. In particular, larger cosine distances between final and initial weights of each layer consistently translate into better generalization performance of the final model. Interestingly, this relation admits a network independent optimum: training procedures during which all layers' weights reach a cosine distance of 1 from their initialization consistently outperform other configurations -by up to 30% test accuracy. Moreover, we show that layer rotations are easily monitored and controlled (helpful for hyperparameter tuning) and potentially provide a unified framework to explain the impact of learning rate tuning, weight decay, learning rate warmups and adaptive gradient methods on generalization and training speed. In an attempt to explain the surprising properties of layer rotation, we show on a 1-layer MLP trained on MNIST that layer rotation correlates with the degree to which features of intermediate layers have been trained.
Rafael Vasquez, Bilal Farooq
2019-04-15
A deep reinforcement learning based multi-objective autonomous braking system is presented. The design of the system is formulated in a continuous action space and seeks to maximize both pedestrian safety and perception as well as passenger comfort. The vehicle agent is trained against a large naturalistic dataset containing pedestrian road-crossing trials in which respondents walked across a road under various traffic conditions within an interactive virtual reality environment. The policy for brake control is learned through computer simulation using two reinforcement learning methods i.e. Proximal Policy Optimization and Deep Deterministic Policy Gradient and the efficiency of each are compared. Results show that the system is able to reduce the negative influence on passenger comfort by half while maintaining safe braking operation.
Bogdan Kulynych, Jamie Hayes, Nikita Samarin, Carmela Troncoso
2018-10-25
Machine-learning models for security-critical applications such as bot, malware, or spam detection, operate in constrained discrete domains. These applications would benefit from having provable guarantees against adversarial examples. The existing literature on provable adversarial robustness of models, however, exclusively focuses on robustness to gradient-based attacks in domains such as images. These attacks model the adversarial cost, e.g., amount of distortion applied to an image, as a
-norm. We argue that this approach is not well-suited to model adversarial costs in constrained domains where not all examples are feasible. We introduce a graphical framework that (1) generalizes existing attacks in discrete domains, (2) can accommodate complex cost functions beyond -norms, including financial cost incurred when attacking a classifier, and (3) efficiently produces valid adversarial examples with guarantees of minimal adversarial cost. These guarantees directly translate into a notion of adversarial robustness that takes into account domain constraints and the adversary's capabilities. We show how our framework can be used to evaluate security by crafting adversarial examples that evade a Twitter-bot detection classifier with provably minimal number of changes; and to build privacy defenses by crafting adversarial examples that evade a privacy-invasive website-fingerprinting classifier.
Chien-Feng Liao, Yu Tsao, Xugang Lu, Hisashi Kawai
2019-04-30
In a noisy environment, a lossy speech signal can be automatically restored by a listener if he/she knows the language well. That is, with the built-in knowledge of a "language model", a listener may effectively suppress noise interference and retrieve the target speech signals. Accordingly, we argue that familiarity with the underlying linguistic content of spoken utterances benefits speech enhancement (SE) in noisy environments. In this study, in addition to the conventional modeling for learning the acoustic noisy-clean speech mapping, an abstract symbolic sequential modeling is incorporated into the SE framework. This symbolic sequential modeling can be regarded as a "linguistic constraint" in learning the acoustic noisy-clean speech mapping function. In this study, the symbolic sequences for acoustic signals are obtained as discrete representations with a Vector Quantized Variational Autoencoder algorithm. The obtained symbols are able to capture high-level phoneme-like content from speech signals. The experimental results demonstrate that the proposed framework can obtain notable performance improvement in terms of perceptual evaluation of speech quality (PESQ) and short-time objective intelligibility (STOI) on the TIMIT dataset.
Melvin Wong, Bilal Farooq
2019-01-18
The emergence of data-driven demand analysis have led to the increased use of generative modelling to learn the probabilistic dependencies between random variables. Although their apparent use has largely been limited to image recognition and classification in recent years, generative machine learning algorithms can be a powerful tool for travel behaviour research by replicating travel behaviour by the underlying properties of data structures. In this paper, we examine the use of generative machine learning approach for analyzing multiple discrete-continuous (MDC) travel behaviour data. We provide a plausible perspective of how we can exploit the use of machine learning techniques to interpret the underlying heterogeneities in the data. We show that generative models are conceptually similar to choice selection behaviour process through information entropy and variational Bayesian inference. Without loss of generality, we consider a restricted Boltzmann machine (RBM) based algorithm with multiple discrete-continuous layer, formulated as a variational Bayesian inference optimization problem. We systematically describe the proposed machine learning algorithm and develop a process of analyzing travel behaviour data from a generative learning perspective. We show parameter stability from model analysis and simulation tests on an open dataset with multiple discrete-continuous dimensions from a data size of 293,330 observations. For interpretability, we derive the conditional probabilities, elasticities and perform statistical analysis on the latent variables. We show that our model can generate statistically similar data distributions for travel forecasting and prediction and performs better than purely discriminative methods in validation. Our results indicate that latent constructs in generative models can accurately represent the joint distribution consistently on MDC data.
