In the realm of agriculture, where crop health is integral to global food security, Our focus is on the early detection of crop diseases. Leveraging Convolutional Neural Networks (CNNs) on a diverse dataset of crop images, our study focuses on the development, training, and optimization of these networks to achieve accurate and timely disease classification. The first segment demonstrates the efficacy of CNN architecture and optimization strategy, showcasing the potential of deep learning models in automating the identification process. The synergy of robust disease detection and interpretability through Explainable Artificial Intelligence (XAI) presented in this work marks a significant stride toward bridging the gap between advanced technology and precision agriculture. By employing visualization, the research seeks to unravel the decision-making processes of our models. XAI Visualization method emerges as notably superior in terms of accuracy, hinting at its better identification of the disease, this method achieves an accuracy of 89.75\%, surpassing both the heat map model and the LIME explanation method. This not only enhances the transparency and trustworthiness of the predictions but also provides invaluable insights for end-users, allowing them to comprehend the diagnostic features considered by the complex algorithm.
Authored by Priyadarshini Patil, Sneha Pamali, Shreya Devagiri, A Sushma, Jyothi Mirje
AI systems face potential hardware security threats. Existing AI systems generally use the heterogeneous architecture of CPU + Intelligent Accelerator, with PCIe bus for communication between them. Security mechanisms are implemented on CPUs based on the hardware security isolation architecture. But the conventional hardware security isolation architecture does not include the intelligent accelerator on the PCIe bus. Therefore, from the perspective of hardware security, data offloaded to the intelligent accelerator face great security risks. In order to effectively integrate intelligent accelerator into the CPU’s security mechanism, a novel hardware security isolation architecture is presented in this paper. The PCIe protocol is extended to be security-aware by adding security information packaging and unpacking logic in the PCIe controller. The hardware resources on the intelligent accelerator are isolated in fine-grained. The resources classified into the secure world can only be controlled and used by the software of CPU’s trusted execution environment. Based on the above hardware security isolation architecture, a security isolation spiking convolutional neural network accelerator is designed and implemented in this paper. The experimental results demonstrate that the proposed security isolation architecture has no overhead on the bandwidth and latency of the PCIe controller. The architecture does not affect the performance of the entire hardware computing process from CPU data offloading, intelligent accelerator computing, to data returning to CPU. With low hardware overhead, this security isolation architecture achieves effective isolation and protection of input data, model, and output data. And this architecture can effectively integrate hardware resources of intelligent accelerator into CPU’s security isolation mechanism.
Authored by Rui Gong, Lei Wang, Wei Shi, Wei Liu, JianFeng Zhang
In this work, we present a comprehensive survey on applications of the most recent transformer architecture based on attention in information security. Our review reveals three primary areas of application: Intrusion detection, Anomaly Detection and Malware Detection. We have presented an overview of attention-based mechanisms and their application in each cybersecurity use case, and discussed open grounds for future trends in Artificial Intelligence enabled information security.
Authored by M. Vubangsi, Sarumi Abidemi, Olukayode Akanni, Auwalu Mubarak, Fadi Al-Turjman
Using Intrusion Detection Systems (IDS) powered by artificial intelligence is presented in the proposed work as a novel method for enhancing residential security. The overarching goal of the study is to design, develop, and evaluate a system that employs artificial intelligence techniques for real-time detection and prevention of unauthorized access in response to the rising demand for such measures. Using anomaly detection, neural networks, and decision trees, which are all examples of machine learning algorithms that benefit from the incorporation of data from multiple sensors, the proposed system guarantees the accurate identification of suspicious activities. Proposed work examines large datasets and compares them to conventional security measures to demonstrate the system s superior performance and prospective impact on reducing home intrusions. Proposed work contributes to the field of residential security by proposing a dependable, adaptable, and intelligent method for protecting homes against the ever-changing types of infiltration threats that exist today.
Authored by Jeneetha J, B.Vishnu Prabha, B. Yasotha, Jaisudha J, C. Senthilkumar, V.Samuthira Pandi
The Internet of Things (IoT) has changed the way we gather medical data in real time. But, it also brings worries about keeping this data safe and private. Ensuring a secure system for IoT is crucial. At the same time, a new technology is emerging that can help the IoT industry a lot. It s called Blockchain technology. It keeps data secure, transparent, and unchangeable. It s like a ledger for tracking lots of connected devices and making them work together. To make IoT even safer, we can use facial recognition with Convolutional Neural Networks (CNN). This paper introduces a healthcare system that combines Blockchain and artificial intelligence in IoT. An implementation of Raspberry Pi E-Health system is presented and evaluated in terms of function s cost. Our system present low cost functions.
Authored by Amina Kessentini, Ibtissem Wali, Mayssa Jarray, Nouri Masmoudi
Alzheimer’s disease (AD) is a disorder that has an impact on the functioning of the brain cells which begins gradually and worsens over time. The early detection of the disease is very crucial as it will increase the chances of benefiting from treatment. There is a possibility for delayed diagnosis of the disease. To overcome this delay, in this work an approach has been proposed using Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) to use active Magnetic Resonance Imaging (MRI) scanned reports of Alzheimer’s patients to classify the stages of AD along with Explainable Artificial Intelligence (XAI) known as Gradient Class Activation Map (Grad-CAM) to highlight the regions of the brain where the disease is detected.
Authored by Savarala Chethana, Sreevathsa Charan, Vemula Srihitha, Suja Palaniswamy, Peeta Pati
Increasing automation in vehicles enabled by increased connectivity to the outside world has exposed vulnerabilities in previously siloed automotive networks like controller area networks (CAN). Attributes of CAN such as broadcast-based communication among electronic control units (ECUs) that lowered deployment costs are now being exploited to carry out active injection attacks like denial of service (DoS), fuzzing, and spoofing attacks. Research literature has proposed multiple supervised machine learning models deployed as Intrusion detection systems (IDSs) to detect such malicious activity; however, these are largely limited to identifying previously known attack vectors. With the ever-increasing complexity of active injection attacks, detecting zero-day (novel) attacks in these networks in real-time (to prevent propagation) becomes a problem of particular interest. This paper presents an unsupervised-learning-based convolutional autoencoder architecture for detecting zero-day attacks, which is trained only on benign (attack-free) CAN messages. We quantise the model using Vitis-AI tools from AMD/Xilinx targeting a resource-constrained Zynq Ultrascale platform as our IDS-ECU system for integration. The proposed model successfully achieves equal or higher classification accuracy (\textgreater 99.5\%) on unseen DoS, fuzzing, and spoofing attacks from a publicly available attack dataset when compared to the state-of-the-art unsupervised learning-based IDSs. Additionally, by cleverly overlapping IDS operation on a window of CAN messages with the reception, the model is able to meet line-rate detection (0.43 ms per window) of high-speed CAN, which when coupled with the low energy consumption per inference, makes this architecture ideally suited for detecting zero-day attacks on critical CAN networks.
Authored by Shashwat Khandelwal, Shanker Shreejith
Attacks against computer system are viewed to be the most serious threat in the modern world. A zero-day vulnerability is an unknown vulnerability to the vendor of the system. Deep learning techniques are widely used for anomaly-based intrusion detection. The technique gives a satisfactory result for known attacks but for zero-day attacks the models give contradictory results. In this work, at first, two separate environments were setup to collect training and test data for zero-day attack. Zero-day attack data were generated by simulating real-time zero-day attacks. Ranking of the features from the train and test data was generated using explainable AI (XAI) interface. From the collected training data more attack data were generated by applying time series generative adversarial network (TGAN) for top 12 features. The train data was concatenated with the AWID dataset. A hybrid deep learning model using Long short-term memory (LSTM) and Convolutional neural network (CNN) was developed to test the zero-day data against the GAN generated concatenated dataset and the original AWID dataset. Finally, it was found that the result using the concatenated dataset gives better performance with 93.53\% accuracy, where the result from only AWID dataset gives 84.29\% accuracy.
Authored by Md. Asaduzzaman, Md. Rahman
In the face of a large number of network attacks, intrusion detection system can issue early warning, indicating the emergence of network attacks. In order to improve the traditional machine learning network intrusion detection model to identify the behavior of network attacks, improve the detection accuracy and accuracy. Convolutional neural network is used to construct intrusion detection model, which has better ability to solve complex problems and better adaptability of algorithm. In order to solve the problems such as dimension explosion caused by input data, the albino PCA algorithm is used to extract data features and reduce data dimensions. For the common problem of convolutional neural networks in intrusion detection such as overfitting, Dropout layers are added before and after the fully connected layer of CNN, and Sigmoid is selected as the intrusion classification prediction function. This reduces the overfitting, improves the robustness of the intrusion detection model, and enhances the fault tolerance and generalization ability of the model to improve the accuracy of the intrusion detection model. The effectiveness of the proposed method in intrusion detection is verified by comparison and analysis of numerical examples.
Authored by Peiqing Zhang, Guangke Tian, Haiying Dong
Device recognition is the primary step toward a secure IoT system. However, the existing equipment recognition technology often faces the problems of unobvious data characteristics and insufficient training samples, resulting in low recognition rate. To address this problem, a convolutional neural network-based IoT device recognition method is proposed. We first extract the background icons of various IoT devices through the Internet, and then use the ResNet50 neural network to extract icon feature vectors to build an IoT icon library, and realize accurate identification of device types through image retrieval. The experimental results show that the accuracy rate of sampling retrieval in the icon library can reach 98.5\%, and the recognition accuracy rate outside the library can reach 83.3\%, which can effectively identify the type of IoT devices.
Authored by Minghao Lu, Linghui Li, Yali Gao, Xiaoyong Li
The network intrusion detection system capably safeguards our network environment from attacks. Yet, the relentless surge in bandwidth and inherent constraints within these systems often hinder detection, particularly in confrontations with substantial traffic volume. Hence, this paper introduces the IP-filtered multi-channel convolutional neural networks (IP-MCCLSTM), which filters traffic by IP, curtails system loading, and notably enhances detection efficiency. IP-MCCLSTM outperforms comparison methods in tests using the 2017CICIDS data set. The result shows IPMCCLSTM obtains 98.9\% accuracy and 99.7\% Macro-Recall rate, showcasing its potential as an avant-garde solution in intrusion detection.
Authored by Qin Feng, Zhang Lin, Liang Bing
In the face of a large number of network attacks, intrusion detection system can issue early warning, indicating the emergence of network attacks. In order to improve the traditional machine learning network intrusion detection model to identify the behavior of network attacks, improve the detection accuracy and accuracy. Convolutional neural network is used to construct intrusion detection model, which has better ability to solve complex problems and better adaptability of algorithm. In order to solve the problems such as dimension explosion caused by input data, the albino PCA algorithm is used to extract data features and reduce data dimensions. For the common problem of convolutional neural networks in intrusion detection such as overfitting, Dropout layers are added before and after the fully connected layer of CNN, and Sigmoid is selected as the intrusion classification prediction function. This reduces the overfitting, improves the robustness of the intrusion detection model, and enhances the fault tolerance and generalization ability of the model to improve the accuracy of the intrusion detection model. The effectiveness of the proposed method in intrusion detection is verified by comparison and analysis of numerical examples.
Authored by Peiqing Zhang, Guangke Tian, Haiying Dong
Alzheimer s disease (AD) is a disorder that has an impact on the functioning of the brain cells which begins gradually and worsens over time. The early detection of the disease is very crucial as it will increase the chances of benefiting from treatment. There is a possibility for delayed diagnosis of the disease. To overcome this delay, in this work an approach has been proposed using Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) to use active Magnetic Resonance Imaging (MRI) scanned reports of Alzheimer s patients to classify the stages of AD along with Explainable Artificial Intelligence (XAI) known as Gradient Class Activation Map (Grad-CAM) to highlight the regions of the brain where the disease is detected.
Authored by Savarala Chethana, Sreevathsa Charan, Vemula Srihitha, Suja Palaniswamy, Peeta Pati
Frequency hopping (FH) technology is one of the most effective technologies in the field of radio countermeasures, meanwhile, the recognition of FH signal has become a research hotspot. FH signal is a typical non-stationary signal whose frequency varies nonlinearly with time and the time-frequency analysis technique provides a very effective method for processing this kind of signal. With the renaissance of deep learning, methods based on time-frequency analysis and deep learning are widely studied. Although these methods have achieved good results, the recognition accuracy still needs to be improved. Through the observation of the datasets, we found that there are still difficult samples that are difficult to identify. Through further analysis, we propose a horizontal spatial attention (HSA) block, which can generate spatial weight vector according to the signal distribution, and then readjust the feature map. The HSA block is a plug-and-play module that can be integrated into common convolutional neural network (CNN) to further improve their performance and these networks with HSA block are collectively called HANets. The HSA block also has the advantages of high recognition accuracy (especially under low SNRs), easy to implant, and almost no influence on the number of parameters. We verified our method on two datasets and a series of comparative experiments show that the proposed method achieves good results on FH datasets.
Authored by Pengcheng Liu, Zhen Han, Zhixin Shi, Meimei Li, Meichen Liu
With the rapid development of underwater sensor networks, the design of underwater demodulators become increasingly significant. However, underwater acoustic communication is faced with many problems such as propagation time delay, multipath effect and Doppler effect due to the complexity of underwater environment. Demodulation of underwater communication signals is a challenging task. To solve this problem, we propose a novel binary phase shift keying (BPSK) demodulator for underwater acoustic communication based on convolutional neural network, which demodulates the modulation data by detecting the position of phase shift. The method proposed in this paper significantly reduces the bit error rate (BER) compared with the results of the traditional method in URPC1 datasets (Underwater Robot Picking Contest).
Authored by Tianshun Han, Zhensheng Shi, Haiyong Zheng, Junyu Dong, Zhaorui Gu, Bing Zheng
Understanding dynamic human behavior based on online video has many applications in security control, crime surveillance, sports, and industrial IoT systems. This paper solves the problem of classifying video data recorded on surveillance cameras in order to identify fragments with instances of shoplifting. It is proposed to use a classifier that is a symbiosis of two neural networks: convolutional and recurrent. The convolutional neural network is used for extraction of features from each frame of the video fragment, and the recurrent network for processing the temporal sequence of processed frames and subsequent classification.
Authored by Lyudmyla Kirichenko, Bohdan Sydorenko, Tamara Radivilova, Petro Zinchenko
Object Oriented Security - Aerial surveillance plays an important role for security applications. It can be further used to monitor borders, restricted zones and critical infrastructure. With the help of drones one can perform surveillance and get the exact location of various objects. Aerial object detection comes with many challenges like the object size which can be as low as 20×20 pixels. Images taken from satellites are hundreds of megapixels. Traditional methods like Histogram of oriented gradients (HOG) and Scale invariant feature transformation (SIFT) were used to extract features from the objects. Then these features were given to machine learning classifier like logistic regression, Support vector machine (SVM) and Random forest (RF) for detection and classification. However, the issue with these methods is that they are highly inaccurate and generated many false detections and misclassifications too. With the evolution of Graphics processing units (GPU) and the introduction of convolutional neural networks (CNN) as well as Deep Learning algorithms situation got changed. Now, it is possible to extract more information and provide better accuracy. In this paper for object detection You only look once version 4 (YOLOv4) is used which is one of the state-of-the-art algorithms. It uses Darknet 53 which is a type of CNN as a backbone for feature extraction. In this work the YOLOv4 based proposed system detect and localize vehicles present in the restricted zone and then geotag them.
Authored by Rohit Jadhav, Rajesh Patil, Akshay Diwan, S. Rathod, Ajay Sharma
Neural Style Transfer - With the emergence of deep perceptual image features, style transfer has become a popular application that repaints a picture while preserving the geometric patterns and textures from a sample image. Our work is devoted to the combination of perceptual features from multiple style images, taken at different scales, e.g. to mix large-scale structures of a style image with fine-scale textures. Surprisingly, this turns out to be difficult, as most deep neural representations are learned to be robust to scale modifications, so that large structures tend to be tangled with smaller scales. Here a multi-scale convolutional architecture is proposed for bi-scale style transfer. Our solution is based on a modular auto-encoder composed of two lightweight modules that are trained independently to transfer style at specific scales, with control over styles and colors.
Authored by Thibault Durand, Julien Rabin, David Tschumperle
Neural Style Transfer - Reducing inter-subject variability between new users and the measured source subjects, and effectively using the information of classification models trained by source subject data, is very important for human–machine interfaces. In this study, we propose a style transfer mapping (STM) and fine-tuning (FT) subject transfer framework using convolutional neural networks (CNNs). To evaluate the performance, we used two types of public surface electromyogram datasets named MyoDatasets and NinaPro database 5. Our proposed framework, STM-FT-CNN, showed the best performances in all cases compared with conventional subject transfer frameworks. In the future, we will build an online processing system that includes this subject transfer framework and verify its performance in online experiments.
Authored by Suguru Kanoga, Takayuki Hoshino, Mitsunori Tada
Neural Style Transfer - Image style transfer is an important research content related to image processing in computer vision. Compared with traditional artificial computing methods, deep learning-based convolutional neural networks in the field of machine learning have powerful advantages. This new method has high computational efficiency and a good style transfer effect. To further improve the quality and efficiency of image style transfer, the pre-trained VGG-16 neural network model and VGG-19 neural network model are used to achieve image style transfer, and the transferred images generated by the two neural networks are compared. The research results show that the use of the VGG-16 convolutional neural network to achieve image style transfer is better and more efficient.
Authored by Yilin Tao
Multiple Fault Diagnosis - Bearings are key transmission parts that are extensively used in rolling mechanical and equipment. Bearing failures can affect the regular running of machines, in serious cases, can cause enormous losses in economy and personnel casualties. Therefore, it is important to implement the research of diagnosing bearing faults. In this paper, a bearing faults diagnosis method was developed based on multiple image inputs and deep convolutional neural network. Firstly, the 1Dvibration signal is transformed into three different types of two-dimensional images: time-frequency image, vibration grayscale image and symmetry dot pattern image, respectively. Enter them into multiple DCNNs separately. Finally, Finally, the nonlinear features of multiple DCNN outputs are fused and classified to achieve bearing fault diagnostics. The experimental results indicate that the diagnosis accuracy of this proposed method is 98.8\%, it can extract the fault features of vibration samples well, and it is an effective bearing fault diagnosis methodology.
Authored by Wei Cui, Guoying Meng, Tingxi Gou, Xingwei Wan
Information Reuse and Security - Successive approximation register analog-to-digital converter (SAR ADC) is widely adopted in the Internet of Things (IoT) systems due to its simple structure and high energy efficiency. Unfortunately, SAR ADC dissipates various and unique power features when it converts different input signals, leading to severe vulnerability to power side-channel attack (PSA). The adversary can accurately derive the input signal by only measuring the power information from the analog supply pin (AVDD), digital supply pin (DVDD), and/or reference pin (Ref) which feed to the trained machine learning models. This paper first presents the detailed mathematical analysis of power side-channel attack (PSA) to SAR ADC, concluding that the power information from AVDD is the most vulnerable to PSA compared with the other supply pin. Then, an LSB-reused protection technique is proposed, which utilizes the characteristic of LSB from the SAR ADC itself to protect against PSA. Lastly, this technique is verified in a 12-bit 5 MS/s secure SAR ADC implemented in 65nm technology. By using the current waveform from AVDD, the adopted convolutional neural network (CNN) algorithms can achieve \textgreater99\% prediction accuracy from LSB to MSB in the SAR ADC without protection. With the proposed protection, the bit-wise accuracy drops to around 50\%.
Authored by Lele Fang, Jiahao Liu, Yan Zhu, Chi-Hang Chan, Rui Martins
In the computer field, cybersecurity has always been the focus of attention. How to detect malware is one of the focuses and difficulties in network security research effectively. Traditional existing malware detection schemes can be mainly divided into two methods categories: database matching and the machine learning method. With the rise of deep learning, more and more deep learning methods are applied in the field of malware detection. Deeper semantic features can be extracted via deep neural network. The main tasks of this paper are as follows: (1) Using machine learning methods and one-dimensional convolutional neural networks to detect malware (2) Propose a machine The method of combining learning and deep learning is used for detection. Machine learning uses LGBM to obtain an accuracy rate of 67.16%, and one-dimensional CNN obtains an accuracy rate of 72.47%. In (2), LGBM is used to screen the importance of features and then use a one-dimensional convolutional neural network, which helps to further improve the detection result has an accuracy rate of 78.64%.
Authored by Da Huo, Xiaoyong Li, Linghui Li, Yali Gao, Ximing Li, Jie Yuan
Aiming at the prevention of information security risk in protection and control of smart substation, a multi-level security defense method of substation based on data aggregation and convolution neural network (CNN) is proposed. Firstly, the intelligent electronic device(IED) uses "digital certificate + digital signature" for the first level of identity authentication, and uses UKey identification code for the second level of physical identity authentication; Secondly, the device group of the monitoring layer judges whether the data report is tampered during transmission according to the registration stage and its own ID information, and the device group aggregates the data using the credential information; Finally, the convolution decomposition technology and depth separable technology are combined, and the time factor is introduced to control the degree of data fusion and the number of input channels of the network, so that the network model can learn the original data and fused data at the same time. Simulation results show that the proposed method can effectively save communication overhead, ensure the reliable transmission of messages under normal and abnormal operation, and effectively improve the security defense ability of smart substation.
Authored by Dong Liu, Yingwei Zhu, Haoliang Du, Lixiang Ruan
In defense and security applications, detection of moving target direction is as important as the target detection and/or target classification. In this study, a methodology for the detection of different mobile targets as approaching or receding was proposed for ground surveillance radar data, and convolutional neural networks (CNN) based on transfer learning were employed for this purpose. In order to improve the classification performance, the use of two key concepts, namely Deep Convolutional Generative Adversarial Network (DCGAN) and decision fusion, has been proposed. With DCGAN, the number of limited available data used for training was increased, thus creating a bigger training dataset with identical distribution to the original data for both moving directions. This generated synthetic data was then used along with the original training data to train three different pre-trained deep convolutional networks. Finally, the classification results obtained from these networks were combined with decision fusion approach. In order to evaluate the performance of the proposed method, publicly available RadEch dataset consisting of eight ground target classes was utilized. Based on the experimental results, it was observed that the combined use of the proposed DCGAN and decision fusion methods increased the detection accuracy of moving target for person, vehicle, group of person and all target groups, by 13.63%, 10.01%, 14.82% and 8.62%, respectively.
Authored by Asli Omeroglu, Hussein Mohammed, Argun Oral, Yucel Ozbek