• Saadgi Pandey
• Samradnyee Pawar
• Sakshi Patil
• Neha Rajput
• Chetan Aher

In today's socio-technological landscape, ensuring women's safety is a pressing concern. This paper presents a deep learning-based surveillance system designed to detect and respond to threats against women in real time. A custom-trained YOLO model is used to recognize harmful actions such as physical assault, weapon possession, inappropriate contact, and fights. To enhance situational awareness, a gender detection CNN assesses the number of males in close proximity to a woman, offering context for potential risk. Additionally, a pose detection module enables real-time fall detection, which may indicate distress or emergency. When a threat or unusual activity is identified, the system sends alerts and activates a built-in chatbot to provide immediate guidance and emotional support. The solution is implemented as a Python Flask web application, complete with secure user authentication features such as login, registration, and password recovery. By integrating advanced deep learning techniques with an accessible web interface, this system offers a practical and intelligent approach to promoting women’s safety in both public and private environments.

Women Safety, Deep Learning, Real-Time Video Analysis, YOLO, Threat Detection, Weapon Detection, Action Recognition, Fall Detection, Pose Estimation, Gender Detection, CNN (Convolutional Neural Network), Flask Web Application, Custom Object Detection, etc.

"Women Safety using Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 6, page no.c137-c142, June-2025, Available :http://www.jetir.org/papers/JETIR2506220.pdf

"Women Safety using Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 6, page no. ppc137-c142, June-2025, Available at : http://www.jetir.org/papers/JETIR2506220.pdf