Real-Time Camera-Based Suspicious Activity Detection for Security Enhancement

This issue proposes the development of a real-time monitoring solution leveraging camera input to detect potential malpractice or suspicious activities, such as unauthorized hand gestures or phone usage, during monitored sessions (e.g., testing environments, secure areas).

**Context & Goals:**
- Enhance security monitoring capabilities within the GSA/https project by introducing an AI-powered video analysis feature.
- Use computer vision libraries (e.g., OpenCV, MediaPipe) to detect:
- Unauthorized hand signals/fingers.
- Phone usage (talking, using phone while present in view).
- Customizable suspicious activity patterns. Face above turning left/right

**Proposed Workflow:**
1. Integrate camera feed capture and processing (OpenCV).
2. Employ pose and gesture recognition for real-time detection.
3. Trigger configurable alerts/logging on detection events.
4. Provide documentation for reproducibility and results.

**Benefits:**
-