OneFlow Flowvision Detection Examples.

Training with coco dataset

./train.sh [/path/to/coco/dataset] [#GPUS] [model_name] [batch_size]

Evaluation with coco dataset

./eval.sh [/path/to/coco/dataset] [#GPUS] [model_name] [batch_size]

model_names = [ssd300_vgg16, fasterrcnn_resnet50_fpn, maskrcnn_resnet50_fpn]

coco dataset

Download coco2017 from links below

2017 Train images [118K/18GB]: http://images.cocodataset.org/zips/train2017.zip
2017 Val images [5K/1GB]: http://images.cocodataset.org/zips/val2017.zip
2017 Train/Val annotations [241MB]: http://images.cocodataset.org/annotations/annotations_trainval2017.zip

unzipping

unzip *.zip

pretraind models

Pretrained model will be automatically downloaded to $HOME/.oneflow/flowvision_cache directory.