The first order derivative of a data density can be estimated efficiently by denoising score matching, and has become an important component in many applications, such as image generation and audio synthesis. Higher order derivatives provide additional local information about the data distribution and enable new applications. Although they can be estimated via automatic differentiation of a learned density model, this can amplify estimation errors and is expensive in high dimensional settings. To overcome these limitations, we propose a method to directly estimate high order derivatives (scores) of a data density from samples. We first show that denoising score matching can be interpreted as a particular case of Tweedie's formula. By leveraging Tweedie's formula on higher order moments, we generalize denoising score matching to estimate higher order derivatives. We demonstrate empirically that models trained with the proposed method can approximate second order derivatives more efficiently and accurately than via automatic differentiation. We show that our models can be used to quantify uncertainty in denoising and to improve the mixing speed of Langevin dynamics via Ozaki discretization for sampling synthetic data and natural images.
Ecole Polytechnique Fédérale de Lausanne（洛桑联邦理工学院）
In this work, we present a new approach to analyze the gradient flow for a positive semi-definite matrix denoising problem in an extensive-rank and high-dimensional regime. We use recent linear pencil techniques of random matrix theory to derive fixed point equations which track the complete time evolution of the matrix-mean-square-error of the problem. The predictions of the resulting fixed point equations are validated by numerical experiments. In this short note we briefly illustrate a few predictions of our formalism by way of examples, and in particular we uncover continuous phase transit…
We consider the problem of estimating a function from $n$ noisy samples whose discrete Total Variation (TV) is bounded by $C_n$. We reveal a deep connection to the seemingly disparate problem of Strongly Adaptive online learning (Daniely et al, 2015) and provide an $O(n \log n)$ time algorithm that attains the near minimax optimal rate of $\tilde O (n^{1/3}C_n^{2/3})$ under squared error loss. The resulting algorithm runs online and optimally adapts to the unknown smoothness parameter $C_n$. This leads to a new and more versatile alternative to wavelets-based methods for (1) adaptively estimat…
UMass Amherst（马萨诸塞大学阿默斯特分校）；National University of Singapore（新加坡国立大学）
Deep learning models leak significant amounts of information about their training datasets. Previous work has investigated training models with differential privacy (DP) guarantees through adding DP noise to the gradients. However, such solutions (specifically, DPSGD), result in large degradations in the accuracy of the trained models. In this paper, we aim at training deep learning models with DP guarantees while preserving model accuracy much better than previous works. Our key technique is to encode gradients to map them to a smaller vector space, therefore enabling us to obtain DP guarante…
Reinforcement learning is a promising approach to learning robotics controllers. It has recently been shown that algorithms based on finite-difference estimates of the policy gradient are competitive with algorithms based on the policy gradient theorem. We propose a theoretical framework for understanding this phenomenon. Our key insight is that many dynamical systems (especially those of interest in robotics control tasks) are nearly deterministic -- i.e., they can be modeled as a deterministic system with a small stochastic perturbation. We show that for such systems, finite-difference estim…
Several methods of estimating the mutual information of random variables have been developed in recent years. They can prove valuable for novel approaches to learning statistically independent features. In this paper, we use one of these methods, a mutual information neural estimation (MINE) network, to present a proof-of-concept of how a neural network can perform linear ICA. We minimize the mutual information, as estimated by a MINE network, between the output units of a differentiable encoder network. This is done by simple alternate optimization of the two networks. The method is shown to …
We implement stacked denoising autoencoders, a class of neural networks that are capable of learning powerful representations of high dimensional data. We describe stochastic gradient descent for unsupervised training of autoencoders, as well as a novel genetic algorithm based approach that makes use of gradient information. We analyze the performance of both optimization algorithms and also the representation learning ability of the autoencoder when it is trained on standard image classification datasets.
University of Amsterdam（阿姆斯特丹大学）；ORTEC（奥泰克）；CIFAR（加拿大高级研究所）
We derive an unbiased estimator for expectations over discrete random variables based on sampling without replacement, which reduces variance as it avoids duplicate samples. We show that our estimator can be derived as the Rao-Blackwellization of three different estimators. Combining our estimator with REINFORCE, we obtain a policy gradient estimator and we reduce its variance using a built-in control variate which is obtained without additional model evaluations. The resulting estimator is closely related to other gradient estimators. Experiments with a toy problem, a categorical Variational …
We introduce a method called TracIn that computes the influence of a training example on a prediction made by the model. The idea is to trace how the loss on the test point changes during the training process whenever the training example of interest was utilized. We provide a scalable implementation of TracIn via: (a) a first-order gradient approximation to the exact computation, (b) saved checkpoints of standard training procedures, and (c) cherry-picking layers of a deep neural network. In contrast with previously proposed methods, TracIn is simple to implement; all it needs is the ability …
Aalto University（阿尔托大学）；Curious AI；Sapienza University of Rome（罗马第一大学）
Trajectory optimization using a learned model of the environment is one of the core elements of model-based reinforcement learning. This procedure often suffers from exploiting inaccuracies of the learned model. We propose to regularize trajectory optimization by means of a denoising autoencoder that is trained on the same trajectories as the model of the environment. We show that the proposed regularization leads to improved planning with both gradient-based and gradient-free optimizers. We also demonstrate that using regularized trajectory optimization leads to rapid initial learning in a se…
Low rank representation of binary matrix is powerful in disentangling sparse individual-attribute associations, and has received wide applications. Existing binary matrix factorization (BMF) or co-clustering (CC) methods often assume i.i.d background noise. However, this assumption could be easily violated in real data, where heterogeneous row- or column-wise probability of binary entries results in disparate element-wise background distribution, and paralyzes the rationality of existing methods. We propose a binary data denoising framework, namely BIND, which optimizes the detection of true p…
Machine Learning models are vulnerable to adversarial attacks that rely on perturbing the input data. This work proposes a novel strategy using Autoencoder Deep Neural Networks to defend a machine learning model against two gradient-based attacks: The Fast Gradient Sign attack and Fast Gradient attack. First we use an autoencoder to denoise the test data, which is trained with both clean and corrupted data. Then, we reduce the dimension of the denoised data using the hidden layer representation of another autoencoder. We perform this experiment for multiple values of the bound of adversarial p…
We introduce a new generative model where samples are produced via Langevin dynamics using gradients of the data distribution estimated with score matching. Because gradients can be ill-defined and hard to estimate when the data resides on low-dimensional manifolds, we perturb the data with different levels of Gaussian noise, and jointly estimate the corresponding scores, i.e., the vector fields of gradients of the perturbed data distribution for all noise levels. For sampling, we propose an annealed Langevin dynamics where we use gradients corresponding to gradually decreasing noise levels as…
The success of deep architectures is at least in part attributed to the layer-by-layer unsupervised pre-training that initializes the network. Various papers have reported extensive empirical analysis focusing on the design and implementation of good pre-training procedures. However, an understanding pertaining to the consistency of parameter estimates, the convergence of learning procedures and the sample size estimates is still unavailable in the literature. In this work, we study pre-training in classical and distributed denoising autoencoders with these goals in mind. We show that the grad…
Gradient-based adversarial attacks on deep neural networks pose a serious threat, since they can be deployed by adding imperceptible perturbations to the test data of any network, and the risk they introduce cannot be assessed through the network's original training performance. Denoising and dimensionality reduction are two distinct methods that have been independently investigated to combat such attacks. While denoising offers the ability to tailor the defense to the specific nature of the attack, dimensionality reduction offers the advantage of potentially removing previously unseen perturb…
CSIRO Data61（澳大利亚联邦科学与工业研究组织数据61研究所）；Australian National University（澳大利亚国立大学）；École Polytechnique Paris Saclay（巴黎萨克雷巴黎综合理工学院）；University of Sydney（悉尼大学）；RIKEN（日本理化学研究所）；University of Tokyo（东京大学）
We introduce a family of pairwise stochastic gradient estimators for gradients of expectations, which are related to the log-derivative trick, but involve pairwise interactions between samples. The simplest example of our new estimator, dubbed the fundamental trick estimator, is shown to arise from either a) introducing and approximating an integral representation based on the fundamental theorem of calculus, or b) applying the reparameterisation trick to an implicit parameterisation under infinitesimal perturbation of the parameters. From the former perspective we generalise to a reproducing …
National Tsing Hua University（国立清华大学）；Mediatek Inc.（联发科股份有限公司）
Many existing conditional score-based data generation methods utilize Bayes' theorem to decompose the gradients of a log posterior density into a mixture of scores. These methods facilitate the training procedure of conditional score models, as a mixture of scores can be separately estimated using a score model and a classifier. However, our analysis indicates that the training objectives for the classifier in these methods may lead to a serious score mismatch issue, which corresponds to the situation that the estimated scores deviate from the true ones. Such an issue causes the samples to be …
Insight Centre for Data Analytics（洞察数据分析中心）；University College Cork（科克大学学院）
We propose denoising dictionary learning (DDL), a simple yet effective technique as a protection measure against adversarial perturbations. We examined denoising dictionary learning on MNIST and CIFAR10 perturbed under two different perturbation techniques, fast gradient sign (FGSM) and jacobian saliency maps (JSMA). We evaluated it against five different deep neural networks (DNN) representing the building blocks of most recent architectures indicating a successive progression of model complexity of each other. We show that each model tends to capture different representations based on their …
Technical University of Munich（慕尼黑工业大学）；University of Southern California（南加州大学）
Convolutional Neural Networks (CNNs) have emerged as highly successful tools for image generation, recovery, and restoration. A major contributing factor to this success is that convolutional networks impose strong prior assumptions about natural images. A surprising experiment that highlights this architectural bias towards natural images is that one can remove noise and corruptions from a natural image without using any training data, by simply fitting (via gradient descent) a randomly initialized, over-parameterized convolutional generator to the corrupted image. While this over-parameteriz…
In this work, we propose \texttt{TimeGrad}, an autoregressive model for multivariate probabilistic time series forecasting which samples from the data distribution at each time step by estimating its gradient. To this end, we use diffusion probabilistic models, a class of latent variable models closely connected to score matching and energy-based methods. Our model learns gradients by optimizing a variational bound on the data likelihood and at inference time converts white noise into a sample of the distribution of interest through a Markov chain using Langevin sampling. We demonstrate experi…
This paper introduces WaveGrad, a conditional model for waveform generation which estimates gradients of the data density. The model is built on prior work on score matching and diffusion probabilistic models. It starts from a Gaussian white noise signal and iteratively refines the signal via a gradient-based sampler conditioned on the mel-spectrogram. WaveGrad offers a natural way to trade inference speed for sample quality by adjusting the number of refinement steps, and bridges the gap between non-autoregressive and autoregressive models in terms of audio quality. We find that it can genera…
Uber AI Labs（Uber AI 实验室）；Georgia Institute of Technology（佐治亚理工学院）；ML Collective
In this paper, we introduce a novel form of value function, $Q(s, s')$, that expresses the utility of transitioning from a state $s$ to a neighboring state $s'$ and then acting optimally thereafter. In order to derive an optimal policy, we develop a forward dynamics model that learns to make next-state predictions that maximize this value. This formulation decouples actions from values while still learning off-policy. We highlight the benefits of this approach in terms of value function transfer, learning within redundant action spaces, and learning off-policy from state observations generated…
In this paper a new dictionary learning algorithm for multidimensional data is proposed. Unlike most conventional dictionary learning methods which are derived for dealing with vectors or matrices, our algorithm, named KTSVD, learns a multidimensional dictionary directly via a novel algebraic approach for tensor factorization as proposed in [3, 12, 13]. Using this approach one can define a tensor-SVD and we propose to extend K-SVD algorithm used for 1-D data to a K-TSVD algorithm for handling 2-D and 3-D data. Our algorithm, based on the idea of sparse coding (using group-sparsity over multidi…
We study estimation of a gradient-sparse parameter vector $\boldsymbolθ^* \in \mathbb{R}^p$, having strong gradient-sparsity $s^*:=\|\nabla_G \boldsymbolθ^*\|_0$ on an underlying graph $G$. Given observations $Z_1,\ldots,Z_n$ and a smooth, convex loss function $\mathcal{L}$ for which $\boldsymbolθ^*$ minimizes the population risk $\mathbb{E}[\mathcal{L}(\boldsymbolθ;Z_1,\ldots,Z_n)]$, we propose to estimate $\boldsymbolθ^*$ by a projected gradient descent algorithm that iteratively and approximately projects gradient steps onto spaces of vectors having small gradient-sparsity over low-degree s…
We consider estimating a piecewise-constant image, or a gradient-sparse signal on a general graph, from noisy linear measurements. We propose and study an iterative algorithm to minimize a penalized least-squares objective, with a penalty given by the "l_0-norm" of the signal's discrete graph gradient. The method proceeds by approximate proximal descent, applying the alpha-expansion procedure to minimize a proximal gradient in each iteration, and using a geometric decay of the penalty parameter across iterations. Under a cut-restricted isometry property for the measurement design, we prove glo…
Queens College, City University of New York（纽约市立大学皇后学院）；University of Michigan（密歇根大学）；Northeastern University（东北大学）
Tensors in the form of multilinear arrays are ubiquitous in data science applications. Captured real-world data, including video, hyperspectral images, and discretized physical systems, naturally occur as tensors and often come with attendant noise. Under the additive noise model and with the assumption that the underlying clean tensor has low rank, many denoising methods have been created that utilize tensor decomposition to effect denoising through low rank tensor approximation. However, all such decomposition methods require estimating the tensor rank, or related measures such as the tensor…
This paper investigates the use of prior computation to estimate the value function to improve sample efficiency in on-policy policy gradient methods in reinforcement learning. Our approach is to estimate the value function from prior computations, such as from the Q-network learned in DQN or the value function trained for different but related environments. In particular, we learn a new value function for the target task while combining it with a value estimate from the prior computation. Finally, the resulting value function is used as a baseline in the policy gradient method. This use of a …
Institute of Infocomm Research, A*Star（新加坡科技研究局资讯通信研究院）；Johannes Kepler University Linz（约翰内斯·开普勒大学林茨分校）
The generative learning phase of Autoencoder (AE) and its successor Denosing Autoencoder (DAE) enhances the flexibility of data stream method in exploiting unlabelled samples. Nonetheless, the feasibility of DAE for data stream analytic deserves in-depth study because it characterizes a fixed network capacity which cannot adapt to rapidly changing environments. An automated construction of a denoising autoeconder, namely deep evolving denoising autoencoder (DEVDAN), is proposed in this paper. DEVDAN features an open structure both in the generative phase and in the discriminative phase where i…
Compressive sensing is an impressive approach for fast MRI. It aims at reconstructing MR image using only a few under-sampled data in k-space, enhancing the efficiency of the data acquisition. In this study, we propose to learn priors based on undecimated wavelet transform and an iterative image reconstruction algorithm. At the stage of prior learning, transformed feature images obtained by undecimated wavelet transform are stacked as an input of denoising autoencoder network (DAE). The highly redundant and multi-scale input enables the correlation of feature images at different channels, whic…
Mjolnir: Breaking the Shield of Perturbation-Protected Gradients via Adaptive Diffusion
The Hong Kong Polytechnic University（香港理工大学）；Wuhan University of Technology（武汉理工大学）；Shenzhen University（深圳大学）；The Hong Kong University of Science and Technology（香港科技大学）
Stochastic neurons can be useful for a number of reasons in deep learning models, but in many cases they pose a challenging problem: how to estimate the gradient of a loss function with respect to the input of such stochastic neurons, i.e., can we "back-propagate" through these stochastic neurons? We examine this question, existing approaches, and present two novel families of solutions, applicable in different settings. In particular, it is demonstrated that a simple biologically plausible formula gives rise to an an unbiased (but noisy) estimator of the gradient with respect to a binary stoc…
Seoul National University（首尔国立大学）；Microsoft Research Asia（微软亚洲研究院）
Denoising diffusion probabilistic models have been recently proposed to generate high-quality samples by estimating the gradient of the data density. The framework defines the prior noise as a standard Gaussian distribution, whereas the corresponding data distribution may be more complicated than the standard Gaussian distribution, which potentially introduces inefficiency in denoising the prior noise into the data sample because of the discrepancy between the data and the prior. In this paper, we propose PriorGrad to improve the efficiency of the conditional diffusion model for speech synthes…
In machine learning, a question of great interest is understanding what examples are challenging for a model to classify. Identifying atypical examples ensures the safe deployment of models, isolates samples that require further human inspection and provides interpretability into model behavior. In this work, we propose Variance of Gradients (VoG) as a valuable and efficient metric to rank data by difficulty and to surface a tractable subset of the most challenging examples for human-in-the-loop auditing. We show that data points with high VoG scores are far more difficult for the model to lea…
Stacked denoising autoencoders (SDAs) have been successfully used to learn new representations for domain adaptation. Recently, they have attained record accuracy on standard benchmark tasks of sentiment analysis across different text domains. SDAs learn robust data representations by reconstruction, recovering original features from data that are artificially corrupted with noise. In this paper, we propose marginalized SDA (mSDA) that addresses two crucial limitations of SDAs: high computational cost and lack of scalability to high-dimensional features. In contrast to SDAs, our approach of mS…
Since convolutional neural networks perform well in learning generalizable image priors from large-scale data, these models have been widely used in image denoising tasks. However, the computational complexity increases dramatically as well on complex model. In this paper, We propose a novel lightweight Complementary Attention Module, which includes a density module and a sparse module, which can cooperatively mine dense and sparse features for feature complementary learning to build an efficient lightweight architecture. Moreover, to reduce the loss of details caused by denoising, this paper …
Mila - Quebec AI Institute（米拉-魁北克人工智能研究所）；University of Montréal（蒙特利尔大学）；Peking University（北京大学）；HEC Montréal（蒙特利尔高等商学院）
We study a fundamental problem in computational chemistry known as molecular conformation generation, trying to predict stable 3D structures from 2D molecular graphs. Existing machine learning approaches usually first predict distances between atoms and then generate a 3D structure satisfying the distances, where noise in predicted distances may induce extra errors during 3D coordinate generation. Inspired by the traditional force field methods for molecular dynamics simulation, in this paper, we propose a novel approach called ConfGF by directly estimating the gradient fields of the log densi…
