Enhancing Barangay Safety: Leveraging Convolutional Neural Networks for Comprehensive Detection of Brawling Incidents in Philippine Barangays

Street Brawling remains a significant concern within the Philippine and global context, posing a threat to community well-being. While extensively recognized as a major issue affecting the younger generation worldwide, the focus on such incidents has predominantly centered around schools. However, in the barangay setting, brawling incidents pose a distinct challenge that demands attention. This paper delves into the realm of action recognition through Machine Learning, aiming to contribute to the proactive identification and understanding of physical violence within the community. Employing Convolutional Neural Networks, the study focuses on training algorithms to detect actions associated with brawling incidents, including punching, and head hitting. The positive outcomes underscore the effectiveness of Convolutional Neural Networks in recognizing these actions. Leveraging CCTV cameras, the findings present an avenue for preventing and addressing brawling incidents within the barangay, ultimately fostering a safer and more secure community environment.