Scale-Free Convolutional Neural Network for Remote Sensing Scene Classification

Emotions expressed by humans can be identified from facial expressions, speech signals, or physiological signals. Among them, the use of physiological signals for emotion classification is a notable emerging area of research. In emotion recognition, a person’s electrocardiogram (ECG) and galvanic skin response (GSR) signals cannot be manipulated, unlike facial and voice signals. Moreover, wearables such as smartwatches and wristbands enable the detection of emotions in people’s naturalistic environment. During the COVID-19 pandemic, it was necessary to detect people’s emotions in order to ensure that appropriate actions were taken according to the prevailing situation and achieve societal balance. Experimentally, the duration of the emotion stimulus period and the social and non-social contexts of participants influence the emotion classification process. Hence, classification of emotions when participants are exposed to the elicitation process for a longer duration and taking into consideration the social context needs to be explored. This work explores the classification of emotions using five pretrained convolutional neural network (CNN) models: MobileNet, NASNetMobile, DenseNet 201, InceptionResnetV2, and EfficientNetB7. The continuous wavelet transform (CWT) coefficients were detected from ECG and GSR recordings from the AMIGOS database with suitable filtering. Scalograms of the sum of frequency coefficients versus time were obtained and converted into images. Emotions were classified using the pre-trained CNN models. The valence and arousal emotion classification accuracy obtained using ECG and GSR data were, respectively, 91.27% and 91.45% using the InceptionResnetV2 CNN classifier and 99.19% and 98.39% using the MobileNet CNN classifier. Other studies have not explored the use of scalograms to represent ECG and GSR CWT features for emotion classification using deep learning models. Additionally, this study provides a novel classification of emotions built on individual and group settings using ECG data. When the participants watched long-duration emotion elicitation videos individually and in groups, the accuracy was around 99.8%. MobileNet had the highest accuracy and shortest execution time. These subject-independent classification methods enable emotion classification independent of varying human behavior.

Understanding the learning mechanism of convolutional neural networks in spectral analysis

Remote sensing image (RSI) scene classification is the foundation and important technology of ground object detection, land use management and geographic analysis. During recent years, convolutional neural networks (CNNs) have achieved significant success and are widely applied in RSI scene classification. However, crafted images that serve as adversarial examples can potentially fool CNNs with high confidence and are hard for human eyes to interpret. For the increasing security and robust requirements of RSI scene classification, the adversarial example problem poses a serious problem for the classification results derived from systems using CNN models, which has not been fully recognized by previous research. In this study, to explore the properties of adversarial examples of RSI scene classification, we create different scenarios by testing two major attack algorithms (i.e., the fast gradient sign method (FGSM) and basic iterative method (BIM)) trained on different RSI benchmark datasets to fool CNNs (i.e., InceptionV1, ResNet and a simple CNN). In the experiment, our results show that CNNs of RSI scene classification are also vulnerable to adversarial examples, and some of them have a fooling rate of over 80%. These adversarial examples are affected by the architecture of CNNs and the type of RSI dataset. InceptionV1 has a fooling rate of less than 5%, which is lower than the others. Adversarial examples generated on the UCM dataset are easier than other datasets. Importantly, we also find that the classes of adversarial examples have an attack selectivity property. Misclassifications of adversarial examples of RSIs are related to the similarity of the original classes in the CNN feature space. Attack selectivity reveals potential classes of adversarial examples and provides insights into the design of defensive algorithms in future research.

EfficientNet convolutional neural networks-based Android malware detection

Scene classification of high resolution remote sensing images plays an important role for a wide range of applications. While significant efforts have been made in developing various methods for scene classification, most of them are based on handcrafted or shallow learning-based features. In this paper, we investigate the use of deep convolutional neural network (CNN) for scene classification. To this end, we first adopt two simple and effective strategies to extract CNN features: (1) using pre-trained CNN models as universal feature extractors, and (2) domain-specifically fine-tuning pre-trained CNN models on our scene classification dataset. Then, scene classification is carried out by using simple classifiers such as linear support vector machine (SVM). In our work, three off-the-shelf CNN models including AlexNet [1], VGGNet [2], and GoogleNet [3] are investigated. Comprehensive evaluations on a publicly available 21 classes land use dataset and comparisons with several state-of-the-art approaches demonstrate that deep CNN features are effective for scene classification of high resolution remote sensing images.

In the realm of computer vision, image classification is a critical issue with many applications, including multimedia content analysis, security and surveillance, and medical imaging. The accuracy of image classification algorithms has considerably increased with the development of deep learning. The discipline of image classification has dramatically benefited from the usage of pre-trained Convolutional Neural Network (CNN) models. In this study, we perform image classification on the Wang dataset composed of 1000 images separated into ten categories, using six different pre-trained CNN models. The research made use of pre-trained versions of VGG16, Densenet, Mobilenet, Inception V3, Resnet50, and Xception models. We assessed the model performances in terms of training and testing accuracy. Testing accuracies for ten unique batches of data were calculated and averaged out. Densenet outperformed state-of-the-art models like Xception by a small margin, mainly due to low data availability. The findings of this study show how pre-trained models have advanced the field of image classification and shed light on how well these models perform in diverse image classification tasks. This study can serve as a valuable reference for researchers and practitioners in the field of computer vision and deep learning and help inform their choices when selecting pre-trained models for image classification, depending on their needs, while also considering data availability and computational constraints.

Remote sensing image scene classification is a challenging task. With the development of deep learning, methods based on convolutional neural networks (CNNs) have made great achievements in remote sensing image scene classification. Since the training of a CNN requires a large number of labeled samples, a generative adversarial network (GAN) for sample generation represents a new opportunity to solve the problem of the limited samples. However, most of the existing GAN-based sample generation methods can only generate unlabeled samples, instead of samples labeled with the corresponding scene category. In this article, to solve the problem, a supervised progressive growing generative adversarial network (SPG-GAN) is proposed for remote sensing image scene classification. The proposed method can generate labeled samples for the remote sensing image scene classification, significantly improving the classification accuracy in the case of limited samples. The SPG-GAN method has two main improvements. First, a conditional generative framework for labeled samples is proposed, in which the label information is added in the channel dimension as the input. By considering the constraints of the label information in the loss function, the network can be trained in the direction of a specific category. As a result, the network can generate remote sensing image scene classification samples with label categories. Second, a progressive growing sample generation method is introduced. In order to ensure that the generated samples have more spatial details, they are generated by progressively adding modules to the generator and discriminator, thereby ensuring that the generated sample is of better quality. After testing on two benchmark datasets and carrying out a large-scale experiment in the central area of the city of Wuhan in China, it was found that the proposed method can obtain a superior scene classification accuracy in the case of limited samples.

This paper proposes a new method, called multilayer stacked covariance pooling (MSCP), for remote sensing scene classification. The innovative contribution of the proposed method is that it is able to naturally combine multilayer feature maps, obtained by pretrained convolutional neural network (CNN) models. Specifically, the proposed MSCP-based classification framework consists of the following three steps. First, a pretrained CNN model is used to extract multilayer feature maps. Then, the feature maps are stacked together, and a covariance matrix is calculated for the stacked features. Each entry of the resulting covariance matrix stands for the covariance of two different feature maps, which provides a natural and innovative way to exploit the complementary information provided by feature maps coming from different layers. Finally, the extracted covariance matrices are used as features for classification by a support vector machine. The experimental results, conducted on three challenging data sets, demonstrate that the proposed MSCP method can not only consistently outperform the corresponding single-layer model but also achieve better classification performance than other pretrained CNN-based scene classification methods.

Aiming at the problems of huge parameters and network degradation caused by simple linear stacked convolution layers or continuous full connection layers in traditional expression recognition methods, two convolution neural network models are designed through depth separation convolution and residual module respectively to widen and deepen the network. Firstly, model A adopts depth separation convolution instead of regular convolution layer, and the global average pooling layer replaces the final full connection layer, utilizes the methods of dropout, batch normalization, activation function of PReLU and image augmentation to avoid over-fitting effectively. Model B adopts pre-trained ResNet50 model to extract facial features, magnifies the images twice by the SRGAN method. Using ensemble method to fuse model A and B, the accuracy is further improved. To verify the feasibility of the method, the model was tested on the FER2013 facial expression dataset, and the performance was compared with the other facial expression recognition algorithms. The final results showed the improved convolutional neural network (CNN) reached the advanced precision of 73.244% in FER2013 dataset, and the experiment data and the number of model parameters all proved the effectiveness of this method.

In deep convolutional neural networks (DCNNs), model size and computation complexity are two important factors governing throughput and energy efficiency when deployed to hardware for inference. Recent works on compact DCNNs as well as pruning methods are effective, yet with drawbacks. For instance, more than half the size of all MobileNet models lies in their last two layers, mainly because compact separable convolution (CONV) layers are not applicable to their last fully connected (FC) layers. Also, in pruning methods, the compression is gained at the expense of irregularity in the DCNN architecture, which necessitates additional indexing memory to address nonzero weights, thereby increasing memory footprint, decompression delays, and energy consumption. In this article, we propose cyclic sparsely connected (CSC) architectures, with memory/computation complexity of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mathcal {O}(N\log {}N)$ </tex-math></inline-formula> , where <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$N$ </tex-math></inline-formula> is the number of nodes/channels given a DCNN layer that, contrary to compact depthwise separable layers, can be used as an overlay for both FC and CONV layers of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mathcal {O}(N^{2})$ </tex-math></inline-formula> . Also, contrary to pruning methods, CSC architectures are structurally sparse and require no indexing due to their cyclic nature. We show that both standard convolution and depthwise convolution layers are special cases of the CSC layers, whose mathematical function, along with FC layers, can be unified into one single formulation and whose hardware implementation can be carried out under one arithmetic logic component. We examine the efficacy of the CSC architectures for compression of LeNet, AlexNet, and MobileNet DCNNs with precision ranging from 2 to 32 bits. More specifically, we surge upon the compact 8-bit quantized 0.5 MobileNet V1 and show that by compressing its last two layers with CSC architectures, the model is compressed by <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\sim 1.5\times $ </tex-math></inline-formula> with a size of only 873 kB and little accuracy loss. Finally, we design a configurable hardware that implements all types of DCNN layers including FC, CONV, depthwise, CSC-FC, and CSC-CONV indistinguishably within a unified pipeline. We implement the hardware on a tiny Xilinx field-programmable gate array (FPGA) for total on-chip processing of the compressed MobileNet that, compared to the related work, has the highest Inference/J while utilizing the smallest FPGA.

Land-cover and land-use classification from aerial images is a challenging problem due to high intra-class diversity and inter-class similarities of the images. To analyze the performances of deep convolutional neural network (CNN) models in this domain, we provide three pre-trained CNN models that are adapted to NWPU-RESISC45 dataset using three different training splits, i.e., 80%, 20%, and 10% ratios. The architectures of all three models are redesigned to be modest in size and their structure is kept simple, yet when tested with the NWPU-RESISC45 dataset, all three models perform comparably to the state-of-the-art models. Each of these models is then used to classify the scenes taken from five well-known datasets in this domain without any fine-tuning. We aim to assess the generalization capabilities of these models on the selected datasets. For better analysis, we considered using top-3 and top-5 accuracies of the models in addition to the best predicted category (top-1) that is usually reported by the models. This way of interpretation is very suitable in this domain, since the datasets contain a high number of fine grained categories with large semantic overlaps. We empirically show that the proposed CNN models actually learn the relevant semantic features in the aerial images better than we observed via standard measures. To the best of our knowledge, this is the first work in this domain that analyzes and presents model generalization performances the way we presented it here.

Visual attention mechanism has been widely used in computer vision and plays a vital role in the research of human action recognition. In this paper, we explore a novel moving target detection mechanism for human action recognition and propose a new 3D CNN (3D Convolutional Neural Network) model, dubbed LaM-2SRN (Local Features Enhanced and Moving target detected 2Stream-ResNet) for extracting and learning attention-enhanced spatiotemporal features. The contributions of this paper are as follow: First, the traditional CAM (Class Activation Map) based visual attention algorithm is used to obtain the optical flow information of the human region, thus eliminating the influence of irrelevant optical flow information (such as background clutter). Second, the ViBe algorithm is used to identify the moving target in the continuous frame, retain the optical flow information of the moving target, and make it complement with the optical flow information of the human region, to obtain a complete motion descriptor. After the motion information is marked in the video frame, we put the marked video frames and the original video frames into the 3D CNN and 2D CNN models respectively, to acquired mixed descriptor, which serves as the basis for video classification. Different from most of the previous target detection algorithms, this paper not only detects the human target, but also detects the moving foreground targets in the video, so that the pre-trained CNN model can obtain more complete motion information from videos. We use RGB-only video data for evaluation on two benchmarks: UCF101 and HMDB51, the experimental results demonstrate that our LaM-2SRN is comparable to the previous state-of-the-art algorithms.

The convolutional layers of standard convolutional neural networks (CNNs) are equivariant to translation. However, the convolution and fully-connected layers are not equivariant or invariant to other affine geometric transformations. Recently, a new class of CNNs is proposed in which the conventional layers of CNNs are replaced with equivariant convolution, pooling, and batch-normalization layers. The final classification layer in equivariant neural networks is invariant to different affine geometric transformations such as rotation, reflection and translation, and the scalar value is obtained by either eliminating the spatial dimensions of filter responses using convolution and down-sampling throughout the network or average is taken over the filter responses. In this work, we propose to integrate the orthogonal moments which gives the high-order statistics of the function as an effective means for encoding global invariance with respect to rotation, reflection and translation in fully-connected layers. As a result, the intermediate layers of the network become equivariant while the classification layer becomes invariant. The most widely used Zernike, pseudo-Zernike and orthogonal Fourier-Mellin moments are considered for this purpose. The effectiveness of the proposed work is evaluated by integrating the invariant transition and fully-connected layer in the architecture of group-equivariant CNNs (G-CNNs) on rotated MNIST and CIFAR10 datasets.

This is the first work to propose generation of the network model of a convolutional neural network (CNN) specifically for power-efficient implementation using the ReRAM technology. State-of-the-art in this area is based on implementation of an already fixed CNN model. It uses parallel crossbar structures to achieve a desired precision quantization of the edge weights using the limited precision provided by individual ReRAM devices. In contrast, in this work we keep the ReRAM crossbar structure in mind during model generation, and target altering a base CNN model such that the resulting implementation will be more power efficient. This is by means of eliminating the parallel crossbars as much as possible, thus reducing the number of required analog-to-digital and digital-to-analog converters which are dominant sources of power consumption. We propose four architectural techniques which are applicable to convolutional and fully-connected layers in a CNN and alter the network model to work with lower precision of the weights with negligible loss in accuracy. In our experiments, we achieve at least 50% power savings with almost no loss in accuracy for popular CNNs compared to ReRAM implementation of the base model.

As an important part of emotion research, facial expression recognition is a necessary condition for intelligent interaction between human and machine, which has an important research significance and a potential commercial value. Convolutional neural network (CNN) is an effective method to recognize facial emotions, which can perform feature extraction and classification simultaneously, and can automatically discover multiple levels of representations in data. Due to the fact that there are millions of parameters involved in training the convolutional neural network model, and there is a large demand for marked samples, transfer learning is often used to fine-tune the pre-trained model for a small target sample set. However, there are often some content differences between data sets during deep transfer learning, which will affect the recognition ability of feature extraction. In order to improve the facial expression recognition ability in transfer learning, a hybrid transfer learning model based on an improved convolution restricted boltzmann machine (CRBM) model and a CNN model is proposed in this paper. This method is fused by two learning abilities of these two models. When the pre-trained CNN model is transferred to a small target set, the CRBM is used to replace the full connection layer in the CNN model, and the CRBM layer and the sofmax layer will be retrained on the target set. The added CRBM layer can not only fully connects all feature maps, but can also learn about the unique statistical characteristics about the target set, which eliminates the influence of content differences between data sets, and extracts higher-order statistical features of facial expression images from the target set. The proposed method is evaluated based on four publicly available facial expression databases: JAFFE, FER2013, SFEW and RAF-DB. The new method can achieve better performance than most state-of-the-art methods, and it can effectively prevent the negative influence of transfer learning features between different data sets.