Human Scream Detection and Analysis for Controlling Crime Rate

This paper affords an innovative framework for actual-time detection and evaluation of human screams to enhance public safety and enhance crime response. Leveraging both machine gaining knowledge of, our machine successfully discriminates scream sounds from ambient environmental noise, even in hard acoustic situations. A sturdy preprocessing pipeline is implemented, which involves loading audio files, padding, log-spectrogram extraction, and min–max normalization to extract reliable spectral features from audio records. these features are then utilized by our custom scream detection model to reap excessive detection accuracy even as minimizing false positives. Inte- grated with a geolocation-based totally alert mechanism, the device automatically sends SMS notifications to nearby regulation enforcement, ensuring timely intervention during capability crime incidents. Experimental reviews show that our approach significantly enhances response times and detection precision, presenting a promising tool for proactive crime prevention.