Face_Recognition_Attendance_System v1.1.0

Face_Recognition_Attendance_System v1.1.0

Stable Windows 11 release with robust webcam/GUI handling, safer encodings, and a reproducible setup.

🚀 Highlights

• Stable GUI on Windows 11 (supports MSMF/DSHOW camera backends)
• Best-match recognition via face_distance + threshold
• One-entry-per-person-per-day CSV logging
• Headless mode (no GUI) for servers/kiosks
• Reproducible environment.yml (Python 3.8, dlib via conda, face_recognition via pip)

📦 Install

• Create env: conda env create -f environment.yml
• Activate: conda activate face_attend
• Verify: python -c "import cv2, dlib, face_recognition, pandas, numpy as np; print('OK')"

▶️ Run

• Start: python Attendance.py
• Optional envs:
  • VIDEO_BACKEND=msmf | dshow
  • DISTANCE_THRESHOLD=0.55
  • DAILY_ROTATION=1
  • NO_GUI=1

🛠️ Fixes

• Correct encodings (single 128D per image; no nested lists)
• Skip bad enrollment images (0 or >1 faces)
• Avoid Qt plugin crashes; add headless fallback
• Defensive frame handling and safe RGB conversion

📌 Notes

• Use 1–3 clear, frontal images per person (one face per image).
• For higher recall, increase DISTANCE_THRESHOLD (e.g., 0.55–0.60).