Skip to content

Latest commit

 

History

History
21 lines (17 loc) · 1.1 KB

README.md

File metadata and controls

21 lines (17 loc) · 1.1 KB
Raw

Before Start

Please follow list to put the BDD100K dataset (train, val, test) in the desired folder. The labels generated by our method are provided here (train and val). We'll call the directory that you cloned ENet-BDD100K-Torch as $ENet_BDD100K_ROOT. Note that this model only uses ENet as backbone, and if you use ENet-SAD, the performance will be better.

Testing

  1. Run test script 
    cd $ENet_BDD100K_ROOT
sh ./experiments/test_ENet.sh
    By now, you should be able to reproduce the result (Accuracy: 0.3656, mIoU: 16.02).

Training

  1. Training ENet model 
    cd $ENet_BDD100K_ROOT
sh ./experiments/train_ENet.sh
    The training process should start and trained models would be saved in experiments/models/ENet-new/ by default.
    Then you can test the trained model following the Testing steps above. If your model position or name is changed, remember to set them to yours accordingly.