This paper presents a normalization mechanism called Instance-Level Meta Normalization (ILM~Norm) to address a learning-to-normalize problem. ILM~Norm learns to predict the normalization parameters via both the feature feed-forward and the gradient back-propagation paths. ILM~Norm provides a meta normalization mechanism and has several good properties. It can be easily plugged into existing instance-level normalization schemes such as Instance Normalization, Layer Normalization, or Group Normalization. ILM~Norm normalizes each instance individually and therefore maintains high performance even when small mini-batch is used. The experimental results show that ILM~Norm well adapts to different network architectures and tasks, and it consistently improves the performance of the original models. The code is available at url{https://github.com/Gasoonjia/ILM-Norm.
National University of Singapore（新加坡国立大学）；Northwestern University（西北大学）；Purdue University（普渡大学）；New Oriental AI Research Academy（新东方人工智能研究院）
Batch Normalization (BN)(Ioffe and Szegedy 2015) normalizes the features of an input image via statistics of a batch of images and hence BN will bring the noise to the gradient of the training loss. Previous works indicate that the noise is important for the optimization and generalization of deep neural networks, but too much noise will harm the performance of networks. In our paper, we offer a new point of view that self-attention mechanism can help to regulate the noise by enhancing instance-specific information to obtain a better regularization effect. Therefore, we propose an attention-ba…
Imperial College London（帝国理工学院）；Shanghai Jiao Tong University（上海交通大学）
Normalization layers are essential in a Deep Convolutional Neural Network (DCNN). Various normalization methods have been proposed. The statistics used to normalize the feature maps can be computed at batch, channel, or instance level. However, in most of existing methods, the normalization for each layer is fixed. Batch-Instance Normalization (BIN) is one of the first proposed methods that combines two different normalization methods and achieve diverse normalization for different layers. However, two potential issues exist in BIN: first, the Clip function is not differentiable at input value…
The University of Texas at El Paso（德克萨斯大学埃尔帕索分校）；Microsoft AI for Good Research Lab（微软益世AI研究实验室）；Georgia Institute of Technology（佐治亚理工学院）；Microsoft Research（微软研究院）
Normalization layers have been shown to improve convergence in deep neural networks, and even add useful inductive biases. In many vision applications the local spatial context of the features is important, but most common normalization schemes including Group Normalization (GN), Instance Normalization (IN), and Layer Normalization (LN) normalize over the entire spatial dimension of a feature. This can wash out important signals and degrade performance. For example, in applications that use satellite imagery, input images can be arbitrarily large; consequently, it is nonsensical to normalize o…
As the deep neural networks are being applied to complex tasks, the size of the networks and architecture increases and their topology becomes more complicated too. At the same time, training becomes slow and at some instances inefficient. This motivated the introduction of various normalization techniques such as Batch Normalization and Layer Normalization. The aforementioned normalization methods use arithmetic operations to compute an approximation statistics (mainly the first and second moments) of the layer's data and use it to normalize it. The aforementioned methods use plain Monte Carl…
University of Michigan（密歇根大学）；Stanford University（斯坦福大学）；NTT Research, Inc.（NTT研究院）
Inspired by BatchNorm, there has been an explosion of normalization layers in deep learning. Recent works have identified a multitude of beneficial properties in BatchNorm to explain its success. However, given the pursuit of alternative normalization layers, these properties need to be generalized so that any given layer's success/failure can be accurately predicted. In this work, we take a first step towards this goal by extending known properties of BatchNorm in randomly initialized deep neural networks (DNNs) to several recently proposed normalization layers. Our primary findings follow: (…
University of Texas at Austin（德克萨斯大学奥斯汀分校）；University of Pennsylvania（宾夕法尼亚大学）；Google Research（谷歌研究院）；Princeton University（普林斯顿大学）；Microsoft Research（微软研究院）
Normalization methods such as batch [Ioffe and Szegedy, 2015], weight [Salimansand Kingma, 2016], instance [Ulyanov et al., 2016], and layer normalization [Baet al., 2016] have been widely used in modern machine learning. Here, we study the weight normalization (WN) method [Salimans and Kingma, 2016] and a variant called reparametrized projected gradient descent (rPGD) for overparametrized least-squares regression. WN and rPGD reparametrize the weights with a scale g and a unit vector w and thus the objective function becomes non-convex. We show that this non-convex formulation has beneficial …
University of California, Berkeley（加州大学伯克利分校）；Google AI（谷歌人工智能）
A significant advance in accelerating neural network training has been the development of normalization methods, permitting the training of deep models both faster and with better accuracy. These advances come with practical challenges: for instance, batch normalization ties the prediction of individual examples with other examples within a batch, resulting in a network that is heavily dependent on batch size. Layer normalization and group normalization are data-dependent and thus must be continually used, even at test-time. To address the issues that arise from using explicit normalization te…
Pohang University of Science and Technology（浦项科技大学）
Recently, various normalization layers have been proposed to stabilize the training of deep neural networks. Among them, group normalization is a generalization of layer normalization and instance normalization by allowing a degree of freedom in the number of groups it uses. However, to determine the optimal number of groups, trial-and-error-based hyperparameter tuning is required, and such experiments are time-consuming. In this study, we discuss a reasonable method for setting the number of groups. First, we find that the number of groups influences the gradient behavior of the group normali…
Normalization layers and activation functions are fundamental components in deep networks and typically co-locate with each other. Here we propose to design them using an automated approach. Instead of designing them separately, we unify them into a single tensor-to-tensor computation graph, and evolve its structure starting from basic mathematical functions. Examples of such mathematical functions are addition, multiplication and statistical moments. The use of low-level mathematical functions, in contrast to the use of high-level modules in mainstream NAS, leads to a highly sparse and large …
We investigate the reasons for the performance degradation incurred with batch-independent normalization. We find that the prototypical techniques of layer normalization and instance normalization both induce the appearance of failure modes in the neural network's pre-activations: (i) layer normalization induces a collapse towards channel-wise constant functions; (ii) instance normalization induces a lack of variability in instance statistics, symptomatic of an alteration of the expressivity. To alleviate failure mode (i) without aggravating failure mode (ii), we introduce the technique "Proxy…
Qualcomm AI Research（高通人工智能研究院）；Qualcomm Technologies, Inc.（高通技术公司）
This technical report describes the details of our TASK1A submission of the DCASE2021 challenge. The goal of the task is to design an audio scene classification system for device-imbalanced datasets under the constraints of model complexity. This report introduces four methods to achieve the goal. First, we propose Residual Normalization, a novel feature normalization method that uses instance normalization with a shortcut path to discard unnecessary device-specific information without losing useful information for classification. Second, we design an efficient architecture, BC-ResNet-Mod, a m…
University of Maryland, College Park（马里兰大学帕克分校）；Facebook（脸书）；Harbin Institute of Technology（哈尔滨工业大学）
We propose a novel deep neural network architecture for the challenging problem of single image dehazing, which aims to recover the clear image from a degraded hazy image. Instead of relying on hand-crafted image priors or explicitly estimating the components of the widely used atmospheric scattering model, our end-to-end system directly generates the clear image from an input hazy image. The proposed network has an encoder-decoder architecture with skip connections and instance normalization. We adopt the convolutional layers of the pre-trained VGG network as encoder to exploit the representa…
Princeton University（普林斯顿大学）；Institute for Advanced Study（高等研究院）
Intriguing empirical evidence exists that deep learning can work well with exoticschedules for varying the learning rate. This paper suggests that the phenomenon may be due to Batch Normalization or BN, which is ubiquitous and provides benefits in optimization and generalization across all standard architectures. The following new results are shown about BN with weight decay and momentum (in other words, the typical use case which was not considered in earlier theoretical analyses of stand-alone BN. 1. Training can be done using SGD with momentum and an exponentially increasing learning rate s…
Computer Vision Laboratory, University of Nottingham, UK（诺丁汉大学计算机视觉实验室）；Samsung AI Center, Cambridge, UK（三星人工智能中心剑桥）；Queen Mary University of London, London, UK（伦敦玛丽女王大学）
This paper addresses the problem of model compression via knowledge distillation. To this end, we propose a new knowledge distillation method based on transferring feature statistics, specifically the channel-wise mean and variance, from the teacher to the student. Our method goes beyond the standard way of enforcing the mean and variance of the student to be similar to those of the teacher through an $L_2$ loss, which we found it to be of limited effectiveness. Specifically, we propose a new loss based on adaptive instance normalization to effectively transfer the feature statistics. The main…
Artistic style transfer aims to transfer the style characteristics of one image onto another image while retaining its content. Existing approaches commonly leverage various normalization techniques, although these face limitations in adequately transferring diverse textures to different spatial locations. Self-Attention-based approaches have tackled this issue with partial success but suffer from unwanted artifacts. Motivated by these observations, this paper aims to combine the best of both worlds: self-attention and normalization. That yields a new plug-and-play module that we name Self-Att…
University of Cambridge（剑桥大学）；Invenia Labs（Invenia实验室）；Microsoft Research（微软研究院）
Modern meta-learning approaches for image classification rely on increasingly deep networks to achieve state-of-the-art performance, making batch normalization an essential component of meta-learning pipelines. However, the hierarchical nature of the meta-learning setting presents several challenges that can render conventional batch normalization ineffective, giving rise to the need to rethink normalization in this setting. We evaluate a range of approaches to batch normalization for meta-learning scenarios, and develop a novel approach that we call TaskNorm. Experiments on fourteen datasets …
Heterogeneous information network (HIN) embedding has recently attracted much attention due to its effectiveness in dealing with the complex heterogeneous data. Meta path, which connects different object types with various semantic meanings, is widely used by existing HIN embedding works. However, several challenges have not been addressed so far. First, different meta paths convey different semantic meanings, while existing works assume that all nodes share same weights for meta paths and ignore the personalized preferences of different nodes on different meta paths. Second, given a meta path…
Existing deep neural networks for histopathology image synthesis cannot generate image styles that align with different organs, and cannot produce accurate boundaries of clustered nuclei. To address these issues, we propose a style-guided instance-adaptive normalization (SIAN) approach to synthesize realistic color distributions and textures for histopathology images from different organs. SIAN contains four phases, semantization, stylization, instantiation, and modulation. The first two phases synthesize image semantics and styles by using semantic maps and learned image style vectors. The in…
Qualcomm AI Research（高通人工智能研究院）；Qualcomm Technologies, Inc.（高通技术公司）
While using two-dimensional convolutional neural networks (2D-CNNs) in image processing, it is possible to manipulate domain information using channel statistics, and instance normalization has been a promising way to get domain-invariant features. However, unlike image processing, we analyze that domain-relevant information in an audio feature is dominant in frequency statistics rather than channel statistics. Motivated by our analysis, we introduce Relaxed Instance Frequency-wise Normalization (RFN): a plug-and-play, explicit normalization module along the frequency axis which can eliminate …
Recently, voice conversion (VC) without parallel data has been successfully adapted to multi-target scenario in which a single model is trained to convert the input voice to many different speakers. However, such model suffers from the limitation that it can only convert the voice to the speakers in the training data, which narrows down the applicable scenario of VC. In this paper, we proposed a novel one-shot VC approach which is able to perform VC by only an example utterance from source and target speaker respectively, and the source and target speaker do not even need to be seen during tra…
Inner-Instance Normalization for Time Series Forecasting
Department of Physics, J.K. Institute of Science（J.K.科学研究院物理系）；School of Science, Harbin Institute of Technology（哈尔滨工业大学科学学院）；School of Science, Great Bay University（大湾大学科学学院）；University of Intelligent Studies（智能研究大学）
IN-Flow: Instance Normalization Flow for Non-stationary Time Series Forecasting
University of Oxford（牛津大学）；Microsoft Research Asia（微软亚洲研究院）；University of Macau（澳门大学）；University of Central Florida（中佛罗里达大学）；State Information Center（国家信息中心）；Tongji University（同济大学）；Arizona State University（亚利桑那州立大学）
RECON: Robust symmetry discovery via Explicit Canonical Orientation Normalization
Department for AI in Society, Science, and Technology, Zuse Institute Berlin (ZIB)（人工智能与社会、科学与技术部门，柏林Zuse研究所）；Cartesia AI；Institute of Mathematics, Technische Universität Berlin（数学研究所，柏林技术大学）
The objective of active learning (AL) is to train classification models with less number of labeled instances by selecting only the most informative instances for labeling. The AL algorithms designed for other data types such as images and text do not perform well on graph-structured data. Although a few heuristics-based AL algorithms have been proposed for graphs, a principled approach is lacking. In this paper, we propose MetAL, an AL approach that selects unlabeled instances that directly improve the future performance of a classification model. For a semi-supervised learning problem, we fo…
Instance-specific algorithm selection (AS) deals with the automatic selection of an algorithm from a fixed set of candidates most suitable for a specific instance of an algorithmic problem class, where "suitability" often refers to an algorithm's runtime. Over the past years, a plethora of algorithm selectors have been proposed. As an algorithm selector is again an algorithm solving a specific problem, the idea of algorithm selection could also be applied to AS algorithms, leading to a meta-AS approach: Given an instance, the goal is to select an algorithm selector, which is then used to selec…
High false-positive rate is a long-standing challenge for anomaly detection algorithms, especially in high-stake applications. To identify the true anomalies, in practice, analysts or domain experts will be employed to investigate the top instances one by one in a ranked list of anomalies identified by an anomaly detection system. This verification procedure generates informative labels that can be leveraged to re-rank the anomalies so as to help the analyst to discover more true anomalies given a time budget. Some re-ranking strategies have been proposed to approximate the above sequential de…
NTT Computer and Data Science Laboratories（NTT计算机与数据科学实验室）；NTT Communication Science Laboratories（NTT通信科学实验室）
The ratio of two probability densities, called a density-ratio, is a vital quantity in machine learning. In particular, a relative density-ratio, which is a bounded extension of the density-ratio, has received much attention due to its stability and has been used in various applications such as outlier detection and dataset comparison. Existing methods for (relative) density-ratio estimation (DRE) require many instances from both densities. However, sufficient instances are often unavailable in practice. In this paper, we propose a meta-learning method for relative DRE, which estimates the rel…
Southwest University（西南大学）；Shandong University（山东大学）；Joint SDU-NTU Centre for Artificial Intelligence Research（山东大学-南洋理工大学人工智能联合研究中心）；George Mason University（乔治梅森大学）
Multi-Instance Multi-Label learning (MIML) models complex objects (bags), each of which is associated with a set of interrelated labels and composed with a set of instances. Current MIML solutions still focus on a single-type of objects and assumes an IID distribution of training data. But these objects are linked with objects of other types, %(i.e., pictures in Facebook link with various users), which also encode the semantics of target objects. In addition, they generally need abundant labeled data for training. To effectively mine interdependent MIML objects of different types, we propose a…
Georgia Institute of Technology（佐治亚理工学院）；Purdue University（普渡大学）
A general framework of unsupervised learning for combinatorial optimization (CO) is to train a neural network (NN) whose output gives a problem solution by directly optimizing the CO objective. Albeit with some advantages over traditional solvers, the current framework optimizes an averaged performance over the distribution of historical problem instances, which misaligns with the actual goal of CO that looks for a good solution to every future encountered instance. With this observation, we propose a new objective of unsupervised learning for CO where the goal of learning is to search for goo…
