Network Architecture Search Research Articles

Detection and classification of damaged wheat kernels based on progressive neural architecture search

Quantity and quality of grain are both closely related to national development and social stability. Grain is lost during storage due to mildew and insects. Detection of damaged grain kernels not only can reduce the loss of grain, but also protect human beings from diseases caused by damaged grain. Therefore, research on the automatic detection of damaged grain is of continued urgency. In this paper, we propose a framework combining spectrogram generative adversarial network and progressive neural architecture search (SPGAN-PNAS) to detect and classify mildew-damaged wheat kernels (MDK), insect-damaged wheat kernels (IDK) and undamaged wheat kernels (UDK). First, the spectrogram generative adversarial network (SPGAN) is designed to enlarge the data set. Second, we apply progressive neural architecture search (PNAS) to generate network structure to classify three types of wheat kernels. An F1 of 96.2% is obtained using the proposed method with 5-fold cross-validation. The results are superior to the classical neural networks for detection and classification of damaged wheat kernels. Experimental results show that the structure of SPGAN-PNAS is feasible and effective.

Learning auto-scale representations for person re-identification

Person re-identification (ReID) is a hot topic in computer vision. The data in the ReID is often collected from the cameras with different views and is affected by other environmental factors, which poses a significant challenge to ReID. The omni-scales proposed by OSNet can extract discriminative feature representations, which shows that omni-scales are adequate for the task of the ReID. However, the OSNet is mainly based on a manually designed network architecture. In the OSnet, each block uses the same architecture and has only four scale feature representations. Inspired by neural network architecture search (NAS), we propose a method of auto-scale representations for ReID. Specifically, we first design the auto-scale block, mainly composed of the Lite 3 × 3 operations and the RCB 1 × 1 operations. The connection status among the Lite 3 × 3 operations is just our search space. Then we give our entire macro network architecture, the auto-scale network, which is mainly composed of 6 auto-scale blocks. Unlike other NAS-related work, each block in our search space does not need to share the same architecture but can maintain a different architecture. In the search process, we propose the entropy regularization, the validity regularization and the consistent regularization to alleviate the discretized gap, no valid path, and meaningless edges, respectively. Finally, we verify the effectiveness of the model we searched on four commonly used datasets. Our model can maintain the same 2.2 M parameters as OSNet but can achieve the performance of SOTA. The mAP on the Market1501 dataset can reach 88.7%.

NAS4RRAM: neural network architecture search for inference on RRAM-based accelerators

NAS4RRAM: neural network architecture search for inference on RRAM-based accelerators

Neural network architecture search with AMBER

Deep learning applied to genomics can learn patterns in biological sequences, but designing such models requires expertise and effort. Recent work demonstrates the efficiency of a neural network architecture search algorithm in optimizing genomic models.

Hybrid evolutionary network architecture search (HyENAS) for convolution class of deep neural networks with applications

AbstractConvolutional Neural Networks (CNNs) and its variants are increasingly used across wide domain of applications achieving high performance measures. For high performance, application specific CNN architecture is required, hence the need for network architecture search (NAS) becomes essential. This paper proposes a hybrid evolutionary approach for network architecture search (HyENAS), and targets convolution class of neural networks. One of the significant contribution of this technique is to completely evolve the high performance network by simultaneously finding network structures and their corresponding parameters. An elegant string representation has been proposed which efficiently represents the network. The concept of sparse block evolving requisite layer wise features for dense network is deployed. This permits the network to dynamically evolve for a specific application. In comparison to the other state‐of‐art methods, the high performance of the proposed HyENAS approach is demonstrated across various benchmark data sets belonging to the domain of malariology, oncology, neurology, ophthalmology, and genomics. Further, to deploy the proposed model on lower hardware specification devices, another salient feature of the HyENAS technique is to seamlessly sift out the simpler network architecture with comparable accuracy.

NAS-Guided Lightweight Multiscale Attention Fusion Network for Hyperspectral Image Classification

Deep learning (DL) has become a hot topic in the research field of hyperspectral image (HSI) classification. However, with increasing depth and size of deep learning methods, its application in mobile and embedded vision applications has brought great challenges. In this article, we address a network architecture search (NAS)-guided lightweight spectral–spatial attention feature fusion network (LMAFN) for HSI classification. The overall architecture of the proposed network is guided by several conclusions of NAS, which achieves fewer parameters and lower computation cost with deeper network structure by exploiting multiscale Ghost grouped with efficient channel attention (ECA) module for adaptively adjusting the weights of different channels. It helps fully extract spectral–spatial discriminant features to avoid information loss of the dimension reduction operation. Specifically, a multilayer feature fusion method is proposed to extract the fusion information of the spectral–spatial features of each layer by considering complementary information of different hierarchical structures. Therefore, high-lever spectral–spatial attributes are gradually exploited along with the increase in layers and the fusion of layers. The experimental verification on three real HSI data sets demonstrates that the proposed framework presents more satisfying classification performance and efficiency with deeper network structure and lower parameter size.

AutoDet: Pyramid Network Architecture Search for Object Detection

Feature pyramids have delivered significant improvement in object detection. However, building effective feature pyramids heavily relies on expert knowledge, and also requires strenuous efforts to balance effectiveness and efficiency. Automatic search methods, such as NAS-FPN, automates the design of feature pyramids, but the low search efficiency makes it difficult to apply in a large search space. In this paper, we propose a novel search framework for a feature pyramid network, called AutoDet, which enables to automatic discovery of informative connections between multi-scale features and configure detection architectures with both high efficiency and state-of-the-art performance. In AutoDet, a new search space is specifically designed for feature pyramids in object detectors, which is more general than NAS-FPN. Furthermore, the architecture search process is formulated as a combinatorial optimization problem and solved by a Simulated Annealing-based Network Architecture Search method (SA-NAS). Compared with existing NAS methods, AutoDet ensures a dramatic reduction in search times. For example, our SA-NAS can be up to 30x faster than reinforcement learning-based approaches. Furthermore, AutoDet is compatible with both one-stage and two-stage structures with all kinds of backbone networks. We demonstrate the effectiveness of AutoDet with outperforming single-model results on the COCO dataset. Without pre-training on OpenImages, AutoDet with the ResNet-101 backbone achieves an AP of 39.7 and 47.3 for one-stage and two-stage architectures, respectively, which surpass current state-of-the-art methods.

SceneNet: Remote sensing scene classification deep learning network using multi-objective neural evolution architecture search

The scene classification approaches using deep learning have been the subject of much attention for remote sensing imagery. However, most deep learning networks have been constructed with a fixed architecture for natural image processing, and they are difficult to apply directly to remote sensing images, due to the more complex geometric structural features. Thus, there is an urgent need for automatic search for the most suitable neural network architecture from the image data in scene classification, in which a powerful search mechanism is required, and the computational complexity and performance error of the searched network should be balanced for a practical choice. In this article, a framework for scene classification network architecture search based on multi-objective neural evolution (SceneNet) is proposed. In SceneNet, the network architecture coding and searching are achieved using an evolutionary algorithm, which can implement a more flexible hierarchical extraction of the remote sensing image scene information. Moreover, the computational complexity and the performance error of the searched network are balanced by employing the multi-objective optimization method, and the competitive neural architectures are obtained in a Pareto solution set. The effectiveness of SceneNet is demonstrated by experimental comparisons with several deep neural networks designed by human experts.

DcaNAS: efficient convolutional network Design for Desktop CPU platforms

The hardware platform is a significant consideration in efficient CNN model design. Most lightweight networks are based on GPUs and mobile devices. However, they are usually not efficient nor fast enough for desktop CPU platforms. In this paper, we aim to explore the design of highly-efficient convolutional architectures for desktop CPU platforms. To achieve our goal, we first derive a series of CNN model design guidelines for CPU-based devices by comparing different computing platforms. Based on these proposed guidelines, we further present a Desktop CPU-Aware network architecture search (DcaNAS) to search for the optimal network structure with lower CPU latency. By combining automatic search and manual design, our DcaNAS achieves better flexibility and efficiency. On the ImageNet benchmark, we employ DcaNAS to produce two CPU-based lightweight CNN models: DcaNAS-L for higher accuracy and DcaNAS-S for faster speed. On a single CPU core, DcaNAS-L achieves 78.8% Top-1 (94.6% Top-5) accuracy at 13.6 FPS (73.5 ms), and our DcaNAS-S achieves extremely low CPU latency (43.1 ms). The results show that our DcaNAS method can obtain new state-of-the-art CPU-based networks.

An accurate tool wear prediction method under different cutting conditions based on network architecture search

Tool wear prediction is of significance in advanced manufacturing industries, as it aims to ensure the quality of parts, improve machining efficiency and reduce machining costs. Existing tool wear monitoring and prediction methods mainly adopt neural network model with fixed architecture, which rely on the researchers’ experience and cannot guarantee accuracy under different cutting conditions. This paper proposes a tool wear prediction method based on network architecture search. which can learn a suitable network structure under different cutting conditions. Experiments shows sufficient improvement in the accuracy of predicting tool wear compared with existing methods.

Balanced Neural Architecture Search and Its Application in Specific Emitter Identification

The performance of a single neural network can vary unexpectedly corresponding to different classification tasks, and thus the network with fixed structure may lack flexibility and often lead to overfitting or underfitting. It is urgent, also the main objective of this paper, to deal with the limitation of the fixed neural network structure on classifying radar signals in different electromagnetic environments. We in this paper propose a variable network architecture search (NAS) mechanism, called balanced-NAS framework , and apply it in specific emitter identification (SEI) to greatly improve the flexibility of model searching. In the proposed balanced-NAS framework, a “block-cell” structure and a controller based recurrent neural network (RNN) are utilized to design models automatically according to external environment. In particular, a balance function is also proposed and utilized in the balanced-NAS framework, acting on the RNN controller to take both the validation accuracy and computational budget into consideration while searching for ideal models. The efficiency of the searching process is further enhanced by exploiting a progressive strategy to design simple and complicate child models where unpromising ones after each evaluation process are obsoleted to release searching space. Simulations and experiments indicate that the proposed balanced-NAS framework is extremely efficient and outperforms the conventional algorithms in classifying radar signals in different environments.

Efficient Fault Diagnosis of Rolling Bearings Using Neural Network Architecture Search and Sharing Weights

Bearing is one of the most vital components of industrial machinery. The failure of bearing causes severe problems in the machinery. Therefore, continuous monitoring for the bearings is essential rather than regular manual checking, with the requirement for accuracy of prediction and efficiency. This paper proposes a novel intelligent bearing fault condition monitoring and diagnosis method focusing on computation efficiency, which is an important aspect of a continuous monitoring and embedded-based diagnosis device. In the proposed method, acoustic emission signals containing bearing health information are converted into 2-D spectrograms by Constant Q-Transform (CQT) before using a convolutional neural network to infer the bearing state. To reduce the latency while maintaining high accuracy, we propose an efficient search space for neural network architecture search, i.e., a channel distribution search, that automatically obtain the best performing network. Moreover, we present a separation between two processes of condition monitoring and fault diagnosis to save overall computing resources, with a policy of sharing weights in the training process and sharing features in the testing process. The experimental results show that the proposed method reduces about 50% inference time compared to previous methods while achieving an accuracy of 99.82% for eight types of single and compound fault diagnosis for variable rotational speeds.

Semi-supervised blockwisely architecture search for efficient lightweight generative adversarial network

In the field of computer vision, methods that use fully supervised learning and fixed deep network structures need to be improved. Currently, many studies are devoted to designing neural architecture search methods to use neural networks in a more flexible way. However, most of these methods use fully supervised learning at the cost of extraordinary GPU training time. In view of the above problems, we propose a semi-supervised generative adversarial network and search network architecture based on block structure. Use real pictures and generated pictures with corresponding real tags and pseudo tags for training, to achieve the purpose of semi-supervised learning. By setting the layer’s hyperparameters to a variable and flexible stacking block structure, network architecture search is achieved. The proposed method realizes image generation and extends to image classification. In the experimental results in Section 4, the training time is greatly reduced and the model performance is improved, which illustrates the efficiency of our method. The code can be found in https://github.com/AICV-CUMT/STASGAN.

Task dependent deep LDA pruning of neural networks

With deep learning’s success, a limited number of popular deep nets have been widely adopted for various vision tasks. However, this usually results in unnecessarily high complexities and possibly many features of low task utility. In this paper, we address this problem by introducing a task-dependent deep pruning framework based on Fisher’s Linear Discriminant Analysis (LDA). The approach can be applied to convolutional, fully-connected, and module-based deep network structures, in all cases leveraging the high decorrelation of neuron motifs found in the pre-decision space and cross-layer deconv dependency. Moreover, we examine our approach’s potential in network architecture search for specific tasks and analyze the influence of our pruning on model robustness to noises and adversarial attacks. Experimental results on datasets of generic objects (ImageNet, CIFAR100) as well as domain specific tasks (Adience, and LFWA) illustrate our framework’s superior performance over state-of-the-art pruning approaches and fixed compact nets (e.g. SqueezeNet, MobileNet). The proposed method successfully maintains comparable accuracies even after discarding most parameters (98%–99% for VGG16, up to 82% for the already compact InceptionNet) and with significant FLOP reductions (83% for VGG16, up to 64% for InceptionNet). Through pruning, we can also derive smaller, but more accurate and more robust models suitable for the task.

NAS-HRIS: Automatic Design and Architecture Search of Neural Network for Semantic Segmentation in Remote Sensing Images.

The segmentation of high-resolution (HR) remote sensing images is very important in modern society, especially in the fields of industry, agriculture and urban modelling. Through the neural network, the machine can effectively and accurately extract the surface feature information. However, using the traditional deep learning methods requires plentiful efforts in order to find a robust architecture. In this paper, we introduce a neural network architecture search (NAS) method, called NAS-HRIS, which can automatically search neural network architecture on the dataset. The proposed method embeds a directed acyclic graph (DAG) into the search space and designs the differentiable searching process, which enables it to learn an end-to-end searching rule by using gradient descent optimization. It uses the Gumbel-Max trick to provide an efficient way when drawing samples from a non-continuous probability distribution, and it improves the efficiency of searching and reduces the memory consumption. Compared with other NAS, NAS-HRIS consumes less GPU memory without reducing the accuracy, which corresponds to a large amount of HR remote sensing imagery data. We have carried out experiments on the WHUBuilding dataset and achieved 90.44% MIoU. In order to fully demonstrate the feasibility of the method, we made a new urban Beijing Building dataset, and conducted experiments on satellite images and non-single source images, achieving better results than SegNet, U-Net and Deeplab v3+ models, while the computational complexity of our network architecture is much smaller.

CALPA-NET: Channel-Pruning-Assisted Deep Residual Network for Steganalysis of Digital Images

Over the past few years, detection performance improvements of deep-learning based steganalyzers have been usually achieved through structure expansion. However, excessive expanded structure results in huge computational cost, storage overheads, and consequently difficulty in training and deployment. In this paper we propose CALPA-NET, a ChAnneL-Pruning-Assisted deep residual network architecture search approach to shrink the network structure of existing vast, over-parameterized deep-learning based steganalyzers. We observe that the broad inverted-pyramid structure of existing deep-learning based steganalyzers might contradict the well-established model diversity oriented philosophy, and therefore is not suitable for steganalysis. Then a hybrid criterion combined with two network pruning schemes is introduced to adaptively shrink every involved convolutional layer in a data-driven manner. The resulting network architecture presents a slender bottleneck-like structure. We have conducted extensive experiments on BOSSBase + BOWS2 dataset, more diverse ALASKA dataset and even a large-scale subset extracted from ImageNet CLS-LOC dataset. The experimental results show that the model structure generated by our proposed CALPA-NET can achieve comparative performance with less than two percent of parameters and about one third FLOPs compared to the original steganalytic model. The new model possesses even better adaptivity, transferability, and scalability.

Grammar Guided Genetic Programming for Network Architecture Search and Road Detection on Aerial Orthophotography

Photogrammetry involves aerial photography of the Earth’s surface and subsequently processing the images to provide a more accurate depiction of the area (Orthophotography). It is used by the Spanish Instituto Geográfico Nacional to update road cartography but requires a significant amount of manual labor due to the need to perform visual inspection of all tiled images. Deep learning techniques (artificial neural networks with more than one hidden layer) can perform road detection but it is still unclear how to find the optimal network architecture. Our main goal is the automatic design of deep neural network architectures with grammar-guided genetic programming. In this kind of evolutive algorithm, all the population individuals (here candidate network architectures) are constrained to rules specified by a grammar that defines valid and useful structural patterns to guide the search process. Grammar used includes well-known complex structures (e.g., Inception-like modules) combined with a custom designed mutation operator (dynamically links the mutation probability to structural diversity). Pilot results show that the system is able to design models for road detection that obtain test accuracies similar to that reached by state-of-the-art models when evaluated over a dataset from the Spanish National Aerial Orthophotography Plan.

SAR Images Change Detection Based on Self-Adaptive Network Architecture

In the last few years, neural networks were introduced to change detection for a better understanding of remote sensing images. However, the designs of these neural networks were time-consuming processes of trial and error, which failed to account for their validity. Thus, a simple and efficient change detection method based on network architecture search in terms of the evolutionary algorithm is proposed to deal with SAR images change detection problems. In the proposed method, an efficient gene encoding is applied to represent the unpredictable optimal depth and the number of neurons in each hidden layer. Besides, a combinatorial evaluation strategy and a self-adaptive network solution selection are designed for effective and reasonable network architectures. What is more, a hidden layer random alignment crossover operator and a drawing lots mutation operator are designed for the enhancement of diversity of network architectures. Experimental results on a few SAR image data sets demonstrate that the proposed method can generate appropriate networks to solving SAR images change detection.

Qualities, Challenges and Future of Genetic Algorithms

Genetic algorithms, computer programs that simulate natural evolution, are increasingly applied across many disciplines. They have been used to solve various optimisation problems from neural network architecture search to strategic games, and to model phenomena of adaptation and learning. Expertise on the qualities and drawbacks of this technique is largely scattered across the literature or former, motivating an compilation of this knowledge at the light of the most recent developments of the field. In this review, we present genetic algorithms, their qualities, limitations and challenges, as well as some future development perspectives. Genetic algorithms are capable of exploring large and complex spaces of possible solutions, to quickly locate promising elements, and provide an adequate modelling tool to describe evolutionary systems, from games to economies. They however suffer from high computation costs, difficult parameter configuration, and crucial representation of the solutions. Recent developments such as GPU, parallel and quantum computing, conception of powerful parameter control methods, and novel approaches in representation strategies, may be keys to overcome those limitations. This compiling review aims at informing practitioners and newcomers in the field alike in their genetic algorithm research, and at outlining promising avenues for future research. It highlights the potential for interdisciplinary research associating genetic algorithms to pulse original discoveries in social sciences, open ended evolution, artificial life and AI.

Efficient network architecture search via multiobjective particle swarm optimization based on decomposition

The efforts devoted to manually increasing the width and depth of convolutional neural network (CNN) usually require a large amount of time and expertise. It has stimulated a rising demand of neural architecture search (NAS) over these years. However, most popular NAS approaches solely optimize for low prediction error without penalizing high structure complexity. To this end, this paper proposes MOPSO/D-Net, a CNN architecture search method with multiobjective particle swarm optimization based on decomposition (MOPSO/D). The main goal is to reformulate NAS as a multiobjective evolutionary optimization problem, where the optimal architecture is learned by minimizing two conflicting objectives, namely the error rate of classification and number of parameters of the network. Along with the hybrid binary encoding and adaptive penalty-based boundary intersection, an improved MOPSO/D is further proposed to solve the formulated multiobjective NAS and provide diverse tradeoff solutions. Experimental studies verify the effectiveness of MOPSO/D-Net compared with current manual and automated CNN generation methods. The proposed algorithm achieves impressive classification performance with a small number of parameters on each of two benchmark datasets, particularly, 0.4% error rate with 0.16M params on MNIST and 5.88% error rate with 8.1M params on CIFAR-10, respectively.