This is a warehouse for DeepLabV3-Xception-pytorch-model, can be used to train your segmentation datasets

Download VOC & COCO & CityScapes： 链接：https://pan.baidu.com/s/1fy0R6NJo0YanfD4n6kgCiw?pwd=0grl 提取码：0grl --来自百度网盘超级会员V3的分享

Before training, you may need to enter the corresponding train_XXX.sh file to modify the batch_size, total number of training times, learning rate and other hyperparameters. Note: When you start training with a shell script, do not directly modify the contents of the train.py file. The parameter size in the shell file should be modified directly. The path of the data set is modified in the mypath.py file.