• Habeebur Rahman. P. H
• Navaneeth Jawahar
• Bushra Sultana. K
• Rajendran. N

Deep Learning is a branch of modern technology and machine learning that is fast being used to solve a multitude of problems that affects one, in daily life. Several social problems can be overcome with the proper implementation and usage of Deep Learning. Rape and harassment cases are unfortunately common problems in a country such as India. Identification of such horrors could go a long way in preventing them. With the advent of powerful processing systems, video analysis can be achieved in real time. The approach suggested in this paper involves devising a harassment detection algorithm that can be deployed in real-time for CCTV cameras, thus, enabling the automatic detection of such a critical situation. This approach is novel in that does not require the victim to do any conscious action such as pressing a SOS button via their phones etc. The system proposed involves determining that a circumstance is indeed critical while it occurs to curb any consequences of delay. It follows the development of a deep learning model trained to identify anomalies in voice, object and movement.

Deep Learning, Harassment detection, real-time, video analysis.

"Physical Harassment Detection using Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 5, page no.54-58, May-2019, Available :http://www.jetir.org/papers/JETIR1905110.pdf

"Physical Harassment Detection using Deep Learning", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 5, page no. pp54-58, May-2019, Available at : http://www.jetir.org/papers/JETIR1905110.pdf