In computer vision and anomaly detection, this research delves into the application of AI-based Deep Anomaly Detectors for the identification of anomalies in images and videos. The escalating growth of digital content necessitates robust and efficient methods for anomaly detection to ensure the integrity and security of visual data. As the volume of visual data continues to surge, conventional anomaly detection methods fall short in addressing the complexities inherent in images and videos. Traditional anomaly detection methods often struggle with the nuanced patterns and variations present in images and videos. The need for a more sophisticated and adaptive approach becomes imperative to identify anomalies accurately amidst the vast and diverse landscape of visual data. This study addresses this gap by leveraging the power of artificial intelligence, specifically Deep Anomaly Detectors, to enhance the accuracy and speed of anomaly detection in visual content. This research aims to bridge this gap by proposing a novel methodology that combines deep learning techniques with anomaly detection to achieve superior results in identifying anomalies in visual content. The proposed methodology involves the utilization of state-of-the-art deep learning architectures, training on a diverse dataset of images and videos to capture intricate patterns associated with anomalies. The model is then fine-tuned to enhance its sensitivity to deviations from normal visual patterns, ensuring a robust anomaly detection system. The results showcase a significant improvement in anomaly detection accuracy compared to traditional methods. The AI-based Deep Anomaly Detector exhibits a high level of sensitivity and specificity, effectively distinguishing anomalies in real-world scenarios, thus validating the efficacy of the proposed method.
M. Elavarasi1, R. Pramodhini2, M. Deshmukh Deepak3, R. Mekala4, Chamandeep Kaur5
Vels Institute of Science, Technology and Advanced Studies, India1, Nitte Meenakshi Institute of Technology, India2, Pravara Rural Engineering College, India3, M. Kumarasamy College of Engineering, India4, Jazan University, Saudi Arabia 5
Anomaly Detection, Deep Learning, Image Analysis, Computer Vision, Video Processing
| January | February | March | April | May | June | July | August | September | October | November | December |
| 26 | 42 | 7 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Published By :
ICTACT
Published In :
ICTACT Journal on Image and Video Processing
( Volume: 14 , Issue: 2 , Pages: 3161 - 3167 )
Date of Publication :
November 2023
Page Views :
313
Full Text Views :
97
|