A Spectrogram Image-Based Network Anomaly Detection System Using Deep Convolutional Neural Network

Khan, Adnan Shahid; Ahmad, Zeeshan; Abdullah, Johari; Ahmad, Farhan

doi:10.1109/access.2021.3088149

Cited by 52 publications

(29 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The literature review highlights the different performances evaluation metrics like Accuracy, Precision, Recall, False alarm rate, True negative rate, and F‐score, that are derived from the confusion matrix. Accuracy : To measure the accuracy of the model (classifier) from confusion matric parameters, it is the number of accurate predictions divided by the total number of predictions and when a dataset is balanced is it a valuable performance indicator (Khan et al, 2021).

Accuracy = \frac{TP + TN}{TP + TN + FP + FN}

Precision : Performance measure for accuracy was good only for the balanced dataset was as for imbalance dataset performance measure is not good, to overcome this precision is used.…”

Section: Performance Evaluation Metricesmentioning

confidence: 99%

Network intrusion detection system: A survey on artificial intelligence‐based techniques

Habeeb

Babu²

2022

Expert Systems

View full text Add to dashboard Cite

High data rate requirements in recent years have resulted in the massive expansion of communication systems, network size and the amount of data generated and processed. This has eventually caused many threats to the communication networks as well due to a more frequent generation of security attacks that are either novel or the mutation of the existing attacks. To secure the networks against such threats, an intrusion detection system (IDS) is considered as one of the promising solutions. The main problem with the IDS is its increased false alarm rate (FAR) in detecting the zero‐day attacks. To improve the detection accuracy and minimizing the FAR, the researchers proposed IDS solutions using artificial intelligence (AI) approaches. In this research, we have systematically reviewed the recent AI‐based network IDS (NIDS) solutions proposed during the period 2016–2021 by the research community. We systematically analysed the proposed NIDS solutions based on their strengths, shortcomings, AI methodology adopted, datasets, and the evaluation metrics used for evaluation purposes. From the review, we observed that the hybrid approach is mostly adopted by the researchers to propose AI‐based NIDS solutions, with a trend shifting to deep learning‐based approaches over the last 2 years. Also, most of the proposed solutions are evaluated using a very old dataset with only a few studies opting for the latest datasets. Finally based on our observations, we highlighted the research challenges and the future research directions to help young researchers to contribute to this field.

show abstract

Accuracy = \frac{TP + TN}{TP + TN + FP + FN}

Precision : Performance measure for accuracy was good only for the balanced dataset was as for imbalance dataset performance measure is not good, to overcome this precision is used.…”

Section: Performance Evaluation Metricesmentioning

confidence: 99%

Network intrusion detection system: A survey on artificial intelligence‐based techniques

Habeeb

Babu²

2022

Expert Systems

View full text Add to dashboard Cite

show abstract

“…The classification process is realized in linear support vector machines by choosing a suitable plane to separate the two classes. Different kernel methods classify the image since a linear plane cannot be drawn in a nonlinear support vector machine [24].…”

Section: Related Studiesmentioning

confidence: 99%

Artificial Intelligence in Foreign Object Classification in Fenceless Robotic Work Cells Using 2-D Safety Cameras

2022

Teh. vjesn.

View full text Add to dashboard Cite

Production systems using robotic manipulators have become common in the last few decades, and the trend is towards fenceless cells that save from space. Thus, the safety and flexibility of these systems have become more critical. The safety systems are based on either sensor data or camera images. Although the flexibility of the camera-based systems is better, conventional image processing methods are sensitive to the working environment. Artificial intelligence may be a powerful tool for them to adapt to change requirements quickly and improve accuracy and stability. In this study, a low-cost 2-D camera-based safety system was designed and installed in an experimental fenceless robotic work cell. The system controller was coupled with three alternative deep learning (ResNet-152, AlexNet, SqueezeNet) and three machine learning modules (support vector machine, random forest and decision tree). These modules were trained using photo images of ten distinct foreign objects penetrating the alarm zone. To include the ever-changing conditions of the industrial environment, disruptive effects including camera vibrations, shadows, reflections, illuminance variations etc. are included by using multiple images up to 550 for each class. Using the restricted data used for training and testing the six systems, the SqueezeNet deep learning model gave the best accuracy of 95% without any over-fitting. Despite this, machine learning-based models have been found to have 100 times faster prediction time than deep learning-based ones. Thus, the safety system can be adapted quickly to any possible changes and noise that may arise from working conditions is prevented, and time losses that may occur in industrial production may be prevented.

show abstract

“…We are expecting that each branch will become an expert at distinguishing sound from a certain string. e characteristics recovered by a common convolutional layer will be used by all six branches [30][31][32]. is can be better understood by looking at the summarized plot of our model given in Figure 3.…”

Section: Model Architecturementioning

confidence: 99%

Transfer Learning for Audio Waveform to Guitar Chord Spectrograms Using the Convolution Neural Network

Jadhav

Patel²,

Jhaveri

et al. 2022

Mobile Information Systems

View full text Add to dashboard Cite

Automatic chord recognition has always been approached as a broad music audition task. The desired output is a succession of time-aligned discrete chord symbols, such as GMaj and Asus2. Automatic music transcription is the process of converting a musical recording into a human-readable and interpretable representation. When dealing with polyphonic sounds or removing certain limits, automatic music transcription remains a difficult undertaking. A guitar, for example, presents a greater challenge, as guitarists can play the same note in a variety of places. The study makes use of CNN functionality to generate the guitar tab; initially, the constant-Q transform was used to turn the input audio file into short time spectrograms that the CNN model utilises to analyse the chord. The paper developed a method for extracting chord sequences and notes from audio recordings of solo guitar performances. For intervals in the supplied audio, the proposed approach outputs chord names and fret-board notes. The model described here has been refined to achieve an accuracy of 88.7%. The model’s ability to properly tag audio clips is an incredible advancement.

show abstract

A Spectrogram Image-Based Network Anomaly Detection System Using Deep Convolutional Neural Network

Cited by 52 publications

References 42 publications

Network intrusion detection system: A survey on artificial intelligence‐based techniques

Network intrusion detection system: A survey on artificial intelligence‐based techniques

Artificial Intelligence in Foreign Object Classification in Fenceless Robotic Work Cells Using 2-D Safety Cameras

Transfer Learning for Audio Waveform to Guitar Chord Spectrograms Using the Convolution Neural Network

Contact Info

Product

Resources

About