fotonet

fotonet is a compact Python object-detection library for local inference, training, evaluation, and model export. The supported model is fotonete.

Install

python -m pip install fotonet

For a development checkout:

git clone https://github.com/hazegreleases/fotonet.git
cd fotonet
python -m pip install -e ".[dev]"

Quick start

Load the verified release checkpoint, then run inference from Python:

from fotonet import Fotonet

model = Fotonet("fotonete")
results = model.predict("image.jpg", conf=0.25, imgsz=640)

for detection in results[0].boxes:
    print(detection.cls, detection.conf, detection.xyxy)

For a BGR frame from OpenCV:

frame = cv2.imread("image.jpg")
results = model.predict_bgr(frame, conf=0.25, imgsz=640)

Train

Training uses a YAML dataset configuration and writes checkpoints to a local run directory:

fotonet train \
  model=fotonete \
  data=path/to/data.yaml \
  epochs=300 \
  batch=16 \
  imgsz=640 \
  run_dir=runs/fotonete

Resume an interrupted run:

fotonet train \
  model=fotonete \
  data=path/to/data.yaml \
  resume=runs/fotonete/fotonet_last.pt

Weights and training outputs are not stored in the Git repository.

Export

from fotonet import Fotonet

model = Fotonet("fotonete")
output = model.export(
    format="onnx",
    path="exports/fotonete.onnx",
    imgsz=640,
)
print(output["artifact"])

See the export guide for available formats and optional dependencies.

Documentation

• Documentation portal
• Installation
• Quick start
• Inference and results
• Training and resume
• Models and runtime
• Model configuration
• Export
• Transform API
• Security
• Contributing

License

Apache License 2.0. See LICENSE.