The original title is "EyeNet: Turning a Webcam Feed Into Actionable Security Incidents" which is 9 words. Let me check the rules:

Short summary

Engineer built EyeNet, a real-time video surveillance system that converts raw footage into actionable security incidents by combining detection, tracking, and severity scoring to eliminate alert fatigue. System tracks objects with ByteTrack, filters false positives through persistence requirements, and routes alerts via SMS/email based on anomaly scores. Code, demo, and full architecture decisions are open-sourced.

• •Real-time video pipeline detects hazards, faces, and uniform violations while filtering false positives through object tracking and persistence thresholds
• •Severity-aware event bus routes alerts based on anomaly scores, cooldowns, and confidence levels to prevent notification spam
• •Open-source implementation with full code, architecture walkthrough, and YouTube demo; built in Python with YOLOv8, ByteTrack, Flask, and SQLite

Generated with AI, which can make mistakes.