Custom_Yolo_Object_Detection

1. Introduction

These 3-files are used for training and testing Yolo Model.

2. Requirements

Python version: 2 or 3

Packages:

1.8.0 <= tensorflow < 2.0.1 (theoretically any version that supports tf.data is ok) opencv-python tqdm

3. Running demos

Please run the below command to test:

python new_test.py --input_image "img.jpg"

4. Training

Use the new_train.py file for training on your own classes and data. The trained checkpoints and weights are also saved for trsting and train_fn() defines the various training and saving scheme and parameters.

5 Various Arguments

The new_args.py file have list of various arguments used while training and testing Yolo model.

6. Reference

For more information, please check https://github.com/wizyoung/YOLOv3_TensorFlow