Face Detection using YOLOv8

This model was fine tuned on a dataset of over 10k images containing human faces. The model was fine tuned for 100 epochs with a batch size of 16 on a single NVIDIA V100 16GB GPU, it took around 140 minutes for the fine tuning to complete.

Downstream Tasks

• Face Detection: This model can directly use this model for face detection or it can be further fine tuned own a custom dataset to improve the prediction capabilities.
• Face Recognition: This model can be fine tuned to for face recognition tasks as well, create a dataset with the images of faces and label them accordingly using name or any ID and then use this model as a base model for fine tuning.

Example Usage

# load libraries
from huggingface_hub import hf_hub_download
from ultralytics import YOLO
from supervision import Detections
from PIL import Image

# download model
model_path = hf_hub_download(repo_id="arnabdhar/YOLOv8-Face-Detection", filename="model.pt")

# load model
model = YOLO(model_path)

# inference
image_path = "/path/to/image"
output = model(Image.open(image_path))
results = Detections.from_ultralytics(output[0])

Links

• Dataset Source: Roboflow Universe
• Weights & Biases: Run Details
• Training Artifacts: training-artifacts