Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{180245,
        author = {Amar Chandrakant Kalukhe and Manish Suryakant Ambuse and Siddharthsingh Dipaksingh Suryawanshi and Mahesh Chandrakant Ravaji and Prof. Chaitali Deshpande},
        title = {SUSPICIOUS ACTIVITY DETECTION IN EXAM USING DEEP LEARNING},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {1},
        pages = {948-951},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=180245},
        abstract = {Suspicious activity detection involves analyzing images or video streams to identify specific joint positions or body part movements in individuals. This project explores how neural networks can be applied to identify unusual or abnormal human behavior from CCTV footage. Over the past fifteen years, detecting such activity has remained a complex issue in the field of computer vision. The relevance of this area is reflected in its many applications, including public surveillance, animal observation, behavior interpretation, sign language detection, hands-free motion tracking, and human-computer interaction enhancement.
Despite advancements in depth-sensing technologies, affordable sensors still face challenges due to their limited suitability for indoor environments and their tendency to produce low-resolution and noisy output. These factors reduce their effectiveness in estimating human posture from depth imagery. To overcome these shortcomings, this research integrates neural network approaches aimed at improving detection efficiency and accuracy. 
Such systems are particularly valuable in monitoring human activity in sensitive or high-risk public locations like airports, schools, banks, bus stands, railway stations, shopping malls, roadways, and parking areas. By supporting efforts to identify potential threats such as theft, violence, accidents, or vandalism, these surveillance tools enhance safety and security. Manual observation alone is insufficient for thorough monitoring, which makes intelligent surveillance systems necessary for evaluating and flagging unusual behaviors. Although there has been significant work done using still images, there remains a noticeable gap in leveraging convolutional neural networks (CNNs) for analyzing video data to detect suspicious actions},
        keywords = {Activity Recognition, Neural Networks, Surveillance Footage, Anomaly Detection.},
        month = {June},
        }