/YOLOv3-to-TensorFlow-Lite-Conversion

Convert your pre-trained YOLOv3 models into its corresponding TensorFlow Lite version and test the resulting TF Lite model.

Primary LanguageJupyter Notebook

YOLOv3 to TensorFlow Lite Conversion

It is assummed that you already have a pre-trained YOLOv3 model setup that needs to be converted into TensorFlow Lite.

Steps:

  1. Clone the repository on your local machine.

  2. Create a folder named YOLOv3_TFLite on your local machine and copy the YOLOv3 model weights i.e. yolov3_training_last.weights and classes.txt files inside it. The directory structure should look something like the following.

YOLOv3_TFLite
|__ yolov3_training_last.weights
|__ classes.txt

  1. Sign in to your Google account and upload the YOLOv3_TFLite folder on Google Drive. Change the access permission to Anyone with the link for both the files inside the YOLOv3_TFLite folder.

  2. Open Google Colab and upload the YOLOv3_to_TFLite_Conversion.ipynb notebook from the repository you had cloned and run the notebook cells one-by-one as instructed. For detailed explain, refer the following document.

  3. Once the YOLOv3 model is converted into its TF Lite version, download the detect.tflite onto your local machine from the YOLOv3_TFLite folder saved on Google Drive. Copy detect.tflite and classes.txt inside the YOLOv3-to-TensorFlow-Lite-Conversion repository you had cloned earlier.

  4. Create a folder named test_images inside the repository and add some images inside it which you would like to test the model on.

  5. Open test_tflite.py file and edit Line 151 by replacing <your_test_image> with the name of image file you want to test.

  6. Run the following command: python test_tflite.py

Output:

GitHub Logo

  • Read the Medium blog for step-by-step implementation.