YOLOv3-tf2.0

YOLOv3 implemented with tensorflow 2.0

how to train on MS COCO 2017

download COCO2017 dataset from here. unzip directory train2017, val2017 and annotations. generate dataset with the following command.

python3 create_dataset.py </path/to/train2017> </path/to/val2017> </path/to/annotations>

upon executing the script successfully, there will be directories named trainset and testset under the root directory of the source code.

then train the model by executing the following command

python3 train_eager.py

or

python3 train_keras.py

save the model from checkpoint generated by eager training with command

python3 save_model_eager.py

save the model from checkpoint generated by keras training with command

python3 save_model_keras.py

here are some results of my model which can be downloaded here with password mvvj. I trained YOLO v3 from scratch for 500k iterations with batch size 8 on a single gtx 1080 ti. which is far from the quality of the official darknet model which was trained on four gtx1080 for 500k iterations. I show my result here anyway.

Loss

Detection results

how to predict with the trained model

detect objects in an image by executing the following command

python3 Predictor.py <path/to/image>