templeblock / mask_rcnn_tf

This is a tensorflow program of Mask_RCNN

Mask_RCNN_tf

This is a tensorflow program of Mask_RCNN

you should create three folders before train: "coco", "checkpoint" and "logs"

put you coco dataset in "coco" folder, it has three sub folder: train2014, val2014 and annotations

the model will be saved in "checkpoint" folder and the training log will be saved in "logs".

demo

put checkpoint files in "checkpoint" folder

run "python demo.py"

it will get image in "images" folder and predict

train

prepare training dataset as readme file in coco

run "python coco_train.py"