search

tiancity-NJU / da-faster-rcnn-PyTorch

An unofficial implementation of 'Domain Adaptive Faster R-CNN for Object Detection in the Wild ’
Other
136 stars 37 forks source link

Please change cityscapes to cityscape in the training command #8

Open gsjoy8888 opened 5 years ago

gsjoy8888 commented 5 years ago

train cityscapes -> cityscapes-foggy

CUDA_VISIBLE_DEVICES=GPU_ID python da_trainval_net.py --dataset cityscapes --net vgg16 --bs 1 --lr 2e-3 --lr_decay_step 6 --cuda

tiancity-NJU commented 5 years ago

got it ! thanks

tiancity-NJU commented 5 years ago

It is my negligence. 'cityscape' has been used for both train & test. You can use the latest script