In this paper we propose a novel approach to model compression termed Architecture Compression. Instead of operating on the weight or filter space of the network like classical model compression methods, our approach operates on the architecture space. A 1-D CNN encoder-decoder is trained to learn a mapping from discrete architecture space to a continuous embedding and back. Additionally, this embedding is jointly trained to regress accuracy and parameter count in order to incorporate information about the architecture's effectiveness on the dataset. During the compression phase, we first encode the network and then perform gradient descent in continuous space to optimize a compression objective function that maximizes accuracy and minimizes parameter count. The final continuous feature is then mapped to a discrete architecture using the decoder. We demonstrate the merits of this approach on visual recognition tasks such as CIFAR-10, CIFAR-100, Fashion-MNIST and SVHN and achieve a greater than 20x compression on CIFAR-10.
Tsinghua University（清华大学）；University of California, Santa Barbara（加州大学圣塔芭芭拉分校）；Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）
As well known, the huge memory and compute costs of both artificial neural networks (ANNs) and spiking neural networks (SNNs) greatly hinder their deployment on edge devices with high efficiency. Model compression has been proposed as a promising technique to improve the running efficiency via parameter and operation reduction. Whereas, this technique is mainly practiced in ANNs rather than SNNs. It is interesting to answer how much an SNN model can be compressed without compromising its functionality, where two challenges should be addressed: i) the accuracy of SNNs is usually sensitive to mo…
Mitsubishi Electric Research Laboratories（三菱电机研究实验室）
Model compression methods are important to allow for easier deployment of deep learning models in compute, memory and energy-constrained environments such as mobile phones. Knowledge distillation is a class of model compression algorithm where knowledge from a large teacher network is transferred to a smaller student network thereby improving the student's performance. In this paper, we show how optimal transport-based loss functions can be used for training a student network which encourages learning student network parameters that help bring the distribution of student features closer to tha…
We propose tensorial neural networks (TNNs), a generalization of existing neural networks by extending tensor operations on low order operands to those on high order ones. The problem of parameter learning is challenging, as it corresponds to hierarchical nonlinear tensor decomposition. We propose to solve the learning problem using stochastic gradient descent by deriving nontrivial backpropagation rules in generalized tensor algebra we introduce. Our proposed TNNs has three advantages over existing neural networks: (1) TNNs naturally apply to high order input object and thus preserve the mult…
Does Compression Imply Generalization? A Minimum Description Length Perspective
Leiden University（莱顿大学）；Centrum Wiskunde & Informatica (CWI)（数学和计算机科学中心）；Hefei University of Technology（合肥工业大学）；Great Bay University（大湾区大学）；Vrije Universiteit Amsterdam（阿姆斯特丹自由大学）；UiT The Arctic University of Norway（挪威北极大学）
Structured pruning is an effective compression technique to reduce the computation of neural networks, which is usually achieved by adding perturbations to reduce network parameters at the cost of slightly increasing training loss. A more reasonable approach is to find a sparse minimizer along the flat minimum valley found by optimizers, i.e. stochastic gradient descent, which keeps the training loss constant. To achieve this goal, we propose the structured directional pruning based on orthogonal projecting the perturbations onto the flat minimum valley. We also propose a fast solver sDprun an…
Highly distributed training of Deep Neural Networks (DNNs) on future compute platforms (offering 100 of TeraOps/s of computational capacity) is expected to be severely communication constrained. To overcome this limitation, new gradient compression techniques are needed that are computationally friendly, applicable to a wide variety of layers seen in Deep Neural Networks and adaptable to variations in network architectures as well as their hyper-parameters. In this paper we introduce a novel technique - the Adaptive Residual Gradient Compression (AdaComp) scheme. AdaComp is based on localized …
Deep Neural Networks (DNNs) are applied in a wide range of usecases. There is an increased demand for deploying DNNs on devices that do not have abundant resources such as memory and computation units. Recently, network compression through a variety of techniques such as pruning and quantization have been proposed to reduce the resource requirement. A key parameter that all existing compression techniques are sensitive to is the compression ratio (e.g., pruning sparsity, quantization bitwidth) of each layer. Traditional solutions treat the compression ratios of each layer as hyper-parameters, …
Most existing studies improve the efficiency of Split learning (SL) by compressing the transmitted features. However, most works focus on dimension-wise compression that transforms high-dimensional features into a low-dimensional space. In this paper, we propose circular convolution-based batch-wise compression for SL (C3-SL) to compress multiple features into one single feature. To avoid information loss while merging multiple features, we exploit the quasi-orthogonality of features in high-dimensional space with circular convolution and superposition. To the best of our knowledge, we are the…
Xidian University（西安电子科技大学）；Guangzhou Institute of Technology, Xidian University（西安电子科技大学广州研究院）
Spiking neural networks (SNNs), as one of the brain-inspired models, has spatio-temporal information processing capability, low power feature, and high biological plausibility. The effective spatio-temporal feature makes it suitable for event streams classification. However, neuromorphic datasets, such as N-MNIST, CIFAR10-DVS, DVS128-gesture, need to aggregate individual events into frames with a new higher temporal resolution for event stream classification, which causes high training and inference latency. In this work, we proposed a spatio-temporal compression method to aggregate individual…
University of Illinois, Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
We describe Swapout, a new stochastic training method, that outperforms ResNets of identical network structure yielding impressive results on CIFAR-10 and CIFAR-100. Swapout samples from a rich set of architectures including dropout, stochastic depth and residual architectures as special cases. When viewed as a regularization method swapout not only inhibits co-adaptation of units in a layer, similar to dropout, but also across network layers. We conjecture that swapout achieves strong regularization by implicitly tying the parameters across layers. When viewed as an ensemble training method, …
Thanks to their state-of-the-art performance, deep neural networks are increasingly used for object recognition. To achieve these results, they use millions of parameters to be trained. However, when targeting embedded applications the size of these models becomes problematic. As a consequence, their usage on smartphones or other resource limited devices is prohibited. In this paper we introduce a novel compression method for deep neural networks that is performed during the learning phase. It consists in adding an extra regularization term to the cost function of fully-connected layers. We co…
The enormous inference cost of deep neural networks can be scaled down by network compression. Pruning is one of the predominant approaches used for deep network compression. However, existing pruning techniques have one or more of the following limitations: 1) Additional energy cost on top of the compute heavy training stage due to pruning and fine-tuning stages, 2) Layer-wise pruning based on the statistics of a particular, ignoring the effect of error propagation in the network, 3) Lack of an efficient estimate for determining the important channels globally, 4) Unstructured pruning require…
Residual networks (ResNets) have recently achieved state-of-the-art on challenging computer vision tasks. We introduce Resnet in Resnet (RiR): a deep dual-stream architecture that generalizes ResNets and standard CNNs and is easily implemented with no computational overhead. RiR consistently improves performance over ResNets, outperforms architectures with similar amounts of augmentation on CIFAR-10, and establishes a new state-of-the-art on CIFAR-100.
In this paper, we investigate lossy compression of deep neural networks (DNNs) by weight quantization and lossless source coding for memory-efficient deployment. Whereas the previous work addressed non-universal scalar quantization and entropy coding of DNN weights, we for the first time introduce universal DNN compression by universal vector quantization and universal source coding. In particular, we examine universal randomized lattice quantization of DNNs, which randomizes DNN weights by uniform random dithering before lattice quantization and can perform near-optimally on any source withou…
1 bit deep neural networks (DNNs), of which both the activations and weights are binarized , are attracting more and more attention due to their high computational efficiency and low memory requirement . However, the drawback of large accuracy dropping also restrict s its application. In this paper, we propose a novel Targeted Acceleration and Compression (TAC) framework to improve the performance of 1 bit deep neural networks W e consider that the acceleration and compression effects of binarizing fully connected layer s are not sufficient to compensate for the accuracy loss caused by it In t…
In this paper we introduce Principal Filter Analysis (PFA), an easy to use and effective method for neural network compression. PFA exploits the correlation between filter responses within network layers to recommend a smaller network that maintain as much as possible the accuracy of the full model. We propose two algorithms: the first allows users to target compression to specific network property, such as number of trainable variable (footprint), and produces a compressed model that satisfies the requested property while preserving the maximum amount of spectral energy in the responses of ea…
The term Neural Architecture Search (NAS) refers to the automatic optimization of network architectures for a new, previously unknown task. Since testing an architecture is computationally very expensive, many optimizers need days or even weeks to find suitable architectures. However, this search time can be significantly reduced if knowledge from previous searches on different tasks is reused. In this work, we propose a generally applicable framework that introduces only minor changes to existing optimizers to leverage this feature. As an example, we select an existing optimizer and demonstra…
Institute of Electronics, National Chiao Tung University（国立交通大学电子研究所）
Deciding the amount of neurons during the design of a deep neural network to maximize performance is not intuitive. In this work, we attempt to search for the neuron (filter) configuration of a fixed network architecture that maximizes accuracy. Using iterative pruning methods as a proxy, we parameterize the change of the neuron (filter) number of each layer with respect to the change in parameters, allowing us to efficiently scale an architecture across arbitrary sizes. We also introduce architecture descent which iteratively refines the parameterized function used for model scaling. The comb…
Zhejiang University（浙江大学）；Stevens Institute of Technology（史蒂文斯理工学院）；Alibaba-Zhejiang University Joint Research Institute of Frontier Technologies（阿里巴巴-浙江大学前沿技术联合研究院）
Knowledge distillation has demonstrated encouraging performances in deep model compression. Most existing approaches, however, require massive labeled data to accomplish the knowledge transfer, making the model compression a cumbersome and costly process. In this paper, we investigate the practical few-shot knowledge distillation scenario, where we assume only a few samples without human annotations are available for each category. To this end, we introduce a principled dual-stage distillation scheme tailored for few-shot data. In the first step, we graft the student blocks one by one onto the…
University of South Australia（南澳大学）；Royal University of Bhutan（不丹皇家大学）；Naresuan University（那黎宣大学）
The goals of this research were to search for Convolutional Neural Network (CNN) architectures, suitable for an on-device processor with limited computing resources, performing at substantially lower Network Architecture Search (NAS) costs. A new algorithm entitled an Early Exit Population Initialisation (EE-PI) for Evolutionary Algorithm (EA) was developed to achieve both goals. The EE-PI reduces the total number of parameters in the search process by filtering the models with fewer parameters than the maximum threshold. It will look for a new model to replace those models with parameters mor…
University College London（伦敦大学学院）；GTN Ltd.（GTN有限公司）；Shanghai Jiao Tong University（上海交通大学）
This paper demonstrates a method for tensorizing neural networks based upon an efficient way of approximating scale invariant quantum states, the Multi-scale Entanglement Renormalization Ansatz (MERA). We employ MERA as a replacement for the fully connected layers in a convolutional neural network and test this implementation on the CIFAR-10 and CIFAR-100 datasets. The proposed method outperforms factorization using tensor trains, providing greater compression for the same level of accuracy and greater accuracy for the same level of compression. We demonstrate MERA layers with 14000 times fewe…
With the development of deep neural networks, the size of network models becomes larger and larger. Model compression has become an urgent need for deploying these network models to mobile or embedded devices. Model quantization is a representative model compression technique. Although a lot of quantization methods have been proposed, many of them suffer from a high quantization error caused by a long-tail distribution of network weights. In this paper, we propose a novel quantization method, called weight normalization based quantization (WNQ), for model compression. WNQ adopts weight normali…
Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）；University of Chinese Academy of Sciences（中国科学院大学）
In recent years, deep neural networks have achieved great success in the field of computer vision. However, it is still a big challenge to deploy these deep models on resource-constrained embedded devices such as mobile robots, smart phones and so on. Therefore, network compression for such platforms is a reasonable solution to reduce memory consumption and computation complexity. In this paper, a novel channel pruning method based on genetic algorithm is proposed to compress very deep Convolution Neural Networks (CNNs). Firstly, a pre-trained CNN model is pruned layer by layer according to th…
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）
Latest algorithms for automatic neural architecture search perform remarkable but are basically directionless in search space and computational expensive in training of every intermediate architecture. In this paper, we propose a method for efficient architecture search called EENA (Efficient Evolution of Neural Architecture). Due to the elaborately designed mutation and crossover operations, the evolution process can be guided by the information have already been learned. Therefore, less computational effort will be required while the searching and training time can be reduced significantly. …
We explore efficient neural architecture search methods and show that a simple yet powerful evolutionary algorithm can discover new architectures with excellent performance. Our approach combines a novel hierarchical genetic representation scheme that imitates the modularized design pattern commonly adopted by human experts, and an expressive search space that supports complex topologies. Our algorithm efficiently discovers architectures that outperform a large number of manually designed models for image classification, obtaining top-1 error of 3.6% on CIFAR-10 and 20.3% when transferred to I…
Modern Convolutional Neural Network (CNN) architectures, despite their superiority in solving various problems, are generally too large to be deployed on resource constrained edge devices. In this paper, we reduce memory usage and floating-point operations required by convolutional layers in CNNs. We compress these layers by generalizing the Kronecker Product Decomposition to apply to multidimensional tensors, leading to the Generalized Kronecker Product Decomposition (GKPD). Our approach yields a plug-and-play module that can be used as a drop-in replacement for any convolutional layer. Exper…
Dataset Color Quantization: A Training-Oriented Framework for Dataset-Level Compression
CFAR, Agency for Science, Technology and Research, Singapore（科技研究局CFAR, 新加坡）；IHPC, Agency for Science, Technology and Research, Singapore（科技研究局IHPC, 新加坡）；National University of Singapore（新加坡国立大学）；University of Electronic Science and Technology of China (UESTC)（电子科技大学）；Nanyang Technological University (NTU), Singapore（南洋理工大学, 新加坡）
Neural networks (NNs) have been successfully deployed in many applications. However, architectural design of these models is still a challenging problem. Moreover, neural networks are known to have a lot of redundancy. This increases the computational cost of inference and poses an obstacle to deployment on Internet-of-Thing sensors and edge devices. To address these challenges, we propose the STEERAGE synthesis methodology. It consists of two complementary approaches: efficient architecture search, and grow-and-prune NN synthesis. The first step, covered in a global search module, uses an acc…
Princeton University（普林斯顿大学）；Google Research（谷歌研究院）；University of Maryland, College Park（马里兰大学帕克分校）
We describe a simple and general neural network weight compression approach, in which the network parameters (weights and biases) are represented in a "latent" space, amounting to a reparameterization. This space is equipped with a learned probability model, which is used to impose an entropy penalty on the parameter representation during training, and to compress the representation using a simple arithmetic coder after training. Classification accuracy and model compressibility is maximized jointly, with the bitrate--accuracy trade-off specified by a hyperparameter. We evaluate the method on …
Recent state-of-the-art methods for neural architecture search (NAS) exploit gradient-based optimization by relaxing the problem into continuous optimization over architectures and shared-weights, a noisy process that remains poorly understood. We argue for the study of single-level empirical risk minimization to understand NAS with weight-sharing, reducing the design of NAS methods to devising optimizers and regularizers that can quickly obtain high-quality solutions to this problem. Invoking the theory of mirror descent, we present a geometry-aware framework that exploits the underlying stru…
Stanford University（斯坦福大学）；University of Massachusetts Amherst（马萨诸塞大学阿默斯特分校）
This paper investigates deep neural network (DNN) compression from the perspective of compactly representing and storing trained parameters. We explore the previously overlooked opportunity of cross-layer architecture-agnostic representation sharing for DNN parameters. To do this, we decouple feedforward parameters from DNN architectures and leverage additive quantization, an extreme lossy compression method invented for image descriptors, to compactly represent the parameters. The representations are then finetuned on task objectives to improve task accuracy. We conduct extensive experiments …
We propose a method to incrementally learn an embedding space over the domain of network architectures, to enable the careful selection of architectures for evaluation during compressed architecture search. Given a teacher network, we search for a compressed network architecture by using Bayesian Optimization (BO) with a kernel function defined over our proposed embedding space to select architectures for evaluation. We demonstrate that our search algorithm can significantly outperform various baseline methods, such as random search and reinforcement learning (Ashok et al., 2018). The compress…
University of Maryland, College Park（马里兰大学帕克分校）；Google Research（谷歌研究院）
We introduce LilNetX, an end-to-end trainable technique for neural networks that enables learning models with specified accuracy-rate-computation trade-off. Prior works approach these problems one at a time and often require post-processing or multistage training which become less practical and do not scale very well for large datasets or architectures. Our method constructs a joint training objective that penalizes the self-information of network parameters in a reparameterized latent space to encourage small model size while also introducing priors to increase structured sparsity in the para…
Southeast University（东南大学）；INSERM（法国国家健康与医学研究院）；Université de Rennes 1（雷恩第一大学）
Although convolutional neural network (CNN) has made great progress, large redundant parameters restrict its deployment on embedded devices, especially mobile devices. The recent compression works are focused on real-value convolutional neural network (Real CNN), however, to our knowledge, there is no attempt for the compression of complex-value convolutional neural network (Complex CNN). Compared with the real-valued network, the complex-value neural network is easier to optimize, generalize, and has better learning potential. This paper extends the commonly used deep compression algorithm fr…
In many learning situations, resources at inference time are significantly more constrained than resources at training time. This paper studies a general paradigm, called Differentiable ARchitecture Compression (DARC), that combines model compression and architecture search to learn models that are resource-efficient at inference time. Given a resource-intensive base architecture, DARC utilizes the training data to learn which sub-components can be replaced by cheaper alternatives. The high-level technique can be applied to any neural architecture, and we report experiments on state-of-the-art…
Deploying deep learning services for time-sensitive and resource-constrained settings such as IoT using edge computing systems is a challenging task that requires dynamic adjustment of inference time. Multi-exit architectures allow deep neural networks to terminate their execution early in order to adhere to tight deadlines at the cost of accuracy. To mitigate this cost, in this paper we introduce a novel method called Multi-Exit Curriculum Learning that utilizes curriculum learning, a training strategy for neural networks that imitates human learning by sorting the training samples based on t…
This project performed image completion on CIFAR-10, a dataset of 60,000 32x32 RGB images, using three different neural network architectures: fully convolutional networks, convolutional networks with fully connected layers, and encoder-decoder convolutional networks. The highest performing model was a deep fully convolutional network, which was able to achieve a mean squared error of .015 when comparing the original image pixel values with the predicted pixel values. As well, this network was able to output in-painted images which appeared real to the human eye.
Neural architectures are the foundation for improving performance of deep neural networks (DNNs). This paper presents deep compositional grammatical architectures which harness the best of two worlds: grammar models and DNNs. The proposed architectures integrate compositionality and reconfigurability of the former and the capability of learning rich features of the latter in a principled way. We utilize AND-OR Grammar (AOG) as network generator in this paper and call the resulting networks AOGNets. An AOGNet consists of a number of stages each of which is composed of a number of AOG building b…
Institute of Computing Technology, Chinese Academy of Sciences（中国科学院计算技术研究所）；University of Chinese Academy of Sciences（中国科学院大学）
Latest algorithms for automatic neural architecture search perform remarkable but few of them can effectively design the number of channels for convolutional neural networks and consume less computational efforts. In this paper, we propose a method for efficient automatic architecture search which is special to the widths of networks instead of the connections of neural architecture. Our method, functionally incremental search based on function-preserving, will explore the number of channels rapidly while controlling the number of parameters of the target network. On CIFAR-10 and CIFAR-100 cla…
School of Computer Science and Information Engineering, Hefei University of Technology（合肥工业大学计算机科学与信息工程学院）；Engineering Research Center of Safety Critical Industrial Measurement and Control Technology, Ministry of Education, Hefei University of Technology（合肥工业大学安全关键工业测量与控制技术教育部工程研究中心）
To apply deep CNNs to mobile terminals and portable devices, many scholars have recently worked on the compressing and accelerating deep convolutional neural networks. Based on this, we propose a novel uniform channel pruning (UCP) method to prune deep CNN, and the modified squeeze-and-excitation blocks (MSEB) is used to measure the importance of the channels in the convolutional layers. The unimportant channels, including convolutional kernels related to them, are pruned directly, which greatly reduces the storage cost and the number of calculations. There are two types of residual blocks in …
Convolution neural network demonstrates great capability for multiple tasks, such as image classification and many others. However, much resource is required to train a network. Hence much effort has been made to accelerate neural network by reducing precision of weights, activation, and gradient. However, these filter-wise quantification methods exist a natural upper limit, caused by the size of the kernel. Meanwhile, with the popularity of small kernel, the natural limit further decrease. To address this issue, we propose a new cross-filter compression method that can provide $\sim32\times$ …
Ocean University of China（中国海洋大学）；Technical University of Munich（慕尼黑工业大学）；Virginia Commonwealth University（弗吉尼亚联邦大学）；University of Leicester（莱斯特大学）；University of Portsmouth（朴茨茅斯大学）
Knowledge distillation has been widely used to produce portable and efficient neural networks which can be well applied on edge devices for computer vision tasks. However, almost all top-performing knowledge distillation methods need to access the original training data, which usually has a huge size and is often unavailable. To tackle this problem, we propose a novel data-free approach in this paper, named Dual Discriminator Adversarial Distillation (DDAD) to distill a neural network without any training data or meta-data. To be specific, we use a generator to create samples through dual disc…
National University of Singapore（新加坡国立大学）；Bytedance Inc.（字节跳动公司）
The recent WSNet [1] is a new model compression method through sampling filterweights from a compact set and has demonstrated to be effective for 1D convolutionneural networks (CNNs). However, the weights sampling strategy of WSNet ishandcrafted and fixed which may severely limit the expression ability of the resultedCNNs and weaken its compression ability. In this work, we present a novel auto-sampling method that is applicable to both 1D and 2D CNNs with significantperformance improvement over WSNet. Specifically, our proposed auto-samplingmethod learns the sampling rules end-to-end instead …
Differentiable architecture search (DARTS) has significantly promoted the development of NAS techniques because of its high search efficiency and effectiveness but suffers from performance collapse. In this paper, we make efforts to alleviate the performance collapse problem for DARTS from two aspects. First, we investigate the expressive power of the supernet in DARTS and then derive a new setup of DARTS paradigm with only training BatchNorm. Second, we theoretically find that random features dilute the auxiliary connection role of skip-connection in supernet optimization and enable search al…
College of Artificial Intelligence, Southwest University（西南大学人工智能学院）；Brain-inspired Computing & Intelligent Control of Chongqing Key Lab（重庆市脑启发计算与智能控制重点实验室）；College of Electronics and Information Engineering, Southwest University（西南大学电子信息工程学院）；Harbin Institute of Technology（哈尔滨工业大学）
Neural network pruning has remarkable performance for reducing the complexity of deep network models. Recent network pruning methods usually focused on removing unimportant or redundant filters in the network. In this paper, by exploring the similarities between feature maps, we propose a novel filter pruning method, Central Filter (CF), which suggests that a filter is approximately equal to a set of other filters after appropriate adjustments. Our method is based on the discovery that the average similarity between feature maps changes very little, regardless of the number of input images. Ba…
