YOLOv8 Multi-Object Detection with Age & Gender Estimation

📌 Overview

This project combines YOLOv8 (You Only Look Once v8) for real-time object detection with deep learning models for Age and Gender classification. It uses a webcam to detect people and estimate their age and gender, while also detecting other objects trained on the COCO dataset.

---

⚡ Features

• Real-time object detection using YOLOv8 (yolov8n.pt).
• Detects persons and applies age & gender prediction.
• Uses OpenCV DNN models (.prototxt + .caffemodel) for classification.
• Displays bounding boxes, confidence scores, and age/gender labels.

---

🛠 Requirements

Install the following dependencies before running:

pip install ultralytics opencv-python torch numpy

Pre-trained Models Required

Download and place the following files in your project directory:

1. yolov8n.pt → Ultralytics YOLOv8
2. deploy.prototxt → Face detection prototxt
3. age_deploy.prototxt and age_net.caffemodel → Age classification model
4. gender_deploy.prototxt and gender_net.caffemodel → Gender classification model

(Age & gender models are usually available from OpenCV’s sample DNN models repository.)

---

▶️ How to Run

1. Clone this project or copy the script.

2. Ensure all .prototxt and .caffemodel files are in the same directory.

3. Run the script:

   python age_gender_yolo.py

4. A webcam window will open showing detections.

5. Press q to exit.

---

🎯 Labels Used

• YOLOv8 → COCO dataset (80 classes, including person)
• Gender → Male, Female
• Age → (0-2), (4-6), (8-12), (15-20), (25-32), (38-43), (48-53), (60-100)

---

📌 Notes

• YOLOv8 detects all objects, but age & gender estimation is only applied when a person is detected.
• You can replace yolov8n.pt with yolov8s.pt for better accuracy (but slightly slower performance).
• Ensure good lighting for more accurate predictions.

---

🚀 Future Improvements

• Improve accuracy using deep learning-based face detectors instead of Caffe models.
• Train custom age/gender datasets for more precise classification.
• Extend project to detect mood/emotions along with age and gender.