/Yolov4-TensorFlow-Lite-Model

Ready to run YOLOv4 implemented in TensorFlow Lite format.

Primary LanguagePython

Yolov4-TensorFlow-Lite-Model

Ready to run YOLOv4 implemented in TensorFlow Lite format.

made-with-python made-with-tensorflow

1. Prepare environment

git clone https://github.com/patryklaskowski/Yolov4-TensorFlow-Lite-Model && \
cd Yolov4-TensorFlow-Lite-Model/ && \
python3.7 -m venv env && \
source env/bin/activate && \
python3.7 -m pip install -U pip && \
python3.7 -m pip install -r requirements.txt

2. Download TensorFlow Lite Model

Go here and download yolov4-coco-416.tflite into ./checkpoints/ directory.

3. Run

flags available:

  • -tf, --tflite: Path to tflite model. Default: './checkpoints/yolov4-coco-416.tflite'.
  • -n, --names: Path to .names file. Default: './data/classes/coco.names'.
  • -s, --size: Input size. Default: 416.
python3.7 run_example.py