EUVIMEDEUVIMED entdecken
Crossref

AI-Assisted Live Violence Detection And Emergency Surveillance System

S Adolphine Shyni, Roshni J P, Ashwini E, R Abishaa, R Yogasri

International Journal of Drug Delivery Technology · 02.06.2026 · 16 (45s)

DOI: 10.25258/ijddt.16.45s.2 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 02.06.2026. Verfasst von S Adolphine Shyni und 4 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The increased rate of illegal activities has fuelled the need to enhance sophisticated surveillance systems that can accurately monitor activities in real-time and respond proactively. Traditional deep learning-based monitoring paradigms often have limited flexibility, delayed notification mechanisms, and inefficient visual and audio signal integration. To address these challenges, the present paper presents an AI-driven electronic mail notification system of active and passive video aggression by compressing the YOLO v12, ResNet-101 based face localization and identity verification, and sound-based distress vocalization analysis into a logical framework. YOLO v12 is applied to detect aggressive behaviours and dangerous weapons (such as firearms and blades) in real-time and offers high accuracy of identification with low computation time. As a method to guarantee identity-oriented hazard identification, the ResNet-101 is used to identify the face and verify it as an identity, with the identified faces matched with a prepared offender database to identify known violators in the monitored area. This permits the platform to recognize and document the existence of verified offenders in instantaneous operation. Moreover, a sound-based vocalization identification element is implemented to label archived inputs to recognize vocalizations as distress-related or non-threatening (joyful), assisting in crisis situations recognition. After detection of aggression, weapons, or confirmed offenders, the site automatically generates electronic mail messages to certified operators who have signed-in using a graphical user interface (GUI). The notification includes timestamps and video sequences recorded to ensure instant circumstantial understanding and quick response.The experimental results reveal that the developed architecture boosts accuracy of identification, response time, and reliability in comparison to the traditional monitoring systems, making it suitable to apply in community locations, transit stations, educational institutions, and personal facilities to ensure effectiveness in crime prevention and surveillance. In emergency situations like hospitals, elderly care centers, rehabilitation facilities, and public medical centers, the proposed system has the potential to detect violent or distress-related activities promptly and efficiently, helping to minimize risks and protect the safety of patients.

Zitieren

APA

Shyni, S. A., P, R. J., E, A., Abishaa, R., & Yogasri, R. (2026). AI-Assisted Live Violence Detection And Emergency Surveillance System. International Journal of Drug Delivery Technology, 16(45s). https://doi.org/10.25258/ijddt.16.45s.2

Vancouver

Shyni SA, P RJ, E A, Abishaa R, Yogasri R. AI-Assisted Live Violence Detection And Emergency Surveillance System. International Journal of Drug Delivery Technology. 2026;16(45s). doi:10.25258/ijddt.16.45s.2

BibTeX
@article{Shyni2026,
  author = {Shyni, S Adolphine and P, Roshni J and E, Ashwini and Abishaa, R and Yogasri, R},
  title = {AI-Assisted Live Violence Detection And Emergency Surveillance System},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {45s},
  doi = {10.25258/ijddt.16.45s.2}
}
Kostenlos starten →