Fork of derronqi/yolov8-face

Differences between original repository and fork:

• Compatibility with PyTorch >=2.5. (🔥)
• Original pretrained models from GitHub releases page. (🔥)
• YOLOv8-Lite-t-Face and YOLOv8-Lite-s-Face compatibility fixes.
• Sample script detect.py for inference.
• Sample script export.py to export model to ONNX format.
• The following deprecations has been fixed:
  • DeprecationWarning: 'np.float' is a deprecated alias for builtin 'float'.
  • FutureWarning: You are using 'torch.load' with 'weights_only=False'.
  • FutureWarning: Cython directive 'language_level' not set.
  • Cython Warning: Using deprecated NumPy API.

Installation

pip install -r requirements.txt

Pretrained models

• Download links:

Method	Model Size (MB)	weights	SHA-256
YOLOv8-Lite-t-Face	5.1 1.2	PyTorch ONNX	bed8216edcff10fb6e3e906ae1ae77d6671b3cc3c810906635d5c76f25fd4b42 971ba0be8f9bbea0f824c52ff68adff88ef3477e22d794c4853250a5fb6dba2d
YOLOv8-Lite-s-Face	15.3 3.7	PyTorch ONNX	82a456c2f307d054e62c5f4898005cf85f7d5a00ad482ccd41b11910f36d04dc 4e6b0b024ff068651f590e2435e46f069dc8cdcfd373c28598843f9408dba462
YOLOv8n-Face	6.1 11.8	PyTorch ONNX	d545bf1add5aa736a4febac4f4f9245a6d596cd0fe70d5d57989fe0cb9e626ca af09683be4937ebe23e9da0346cbf46cd937b73ed5fac986e6205e6c7cdd2c25

• Evaluation results on WIDERFace dataset:

Method	Test Size	Easy	Medium	Hard
YOLOv8-Lite-t-Face	640	90.4	87.7	73.3
YOLOv8-Lite-s-Face	640	93.4	91.2	78.6
YOLOv8n-Face	640	94.5	92.2	79.0

Inference

python detect.py --weights yolov8n-face-derronqi.pt --source ultralytics/assets/bus.jpg --save-img

WIDERFace evaluation

• Download WIDERFace validation dataset.
• Move dataset to data/widerface/val directory.

python test_widerface.py --weights yolov8n-face-derronqi.pt --dataset_folder data/widerface/val/images/

cd widerface_evaluate

python setup.py build_ext --inplace

python evaluation.py

Export to ONNX format

pip install onnx

python export.py --weights yolov8n-face-derronqi.pt