deleted: lanenet/model/BiseNet.py

README.md

@@ -1,3 +1,20 @@
 # pytorch-lanenet
 效果很好的lanenet网络，主干网络基于bisenetv2并对主干网络做了修改，效果远好于bisnetv2
+
 可直接训练自己的数据应用于生产
+
+inspired by https://github.com/MaybeShewill-CV/lanenet-lane-detection and https://github.com/leonfrank/lanenet-danet-pytorch
+
+Using Bisenetv2 as Encoder.
+
+install(安装)
+
+
+python setup.py install
+
+Usage(使用)
+
+Train(训练)
+
+
+

lanenet/config.py

@@ -1,8 +1,30 @@
 # coding: utf-8
+# number of unique pixel values
 n_labels=6
+#embeding特征数
 no_of_instances=6
+#分割目标的种类数 只分割车道线值为2
 num_classes=2
+#device_ids 中第一个gpu号
 gpu_no='cuda:0'
+# 使用的gpu
 device_ids = [0, 1,2,3,4,5,6]
+lr=0.001
+epochs=2000
+bs=16
+show_interval=30
+save_interval=3
 # 1-train，2-val
-is_training=1
+is_training=1
+# 模型保存位置
+save_path='./checkpoints'
+#loss权重设置详见compute_loss
+w1=0.25
+#loss权重设置详见compute_loss
+w2=0.25
+#loss权重设置详见compute_loss
+w3=0.25
+#loss权重设置详见compute_loss
+w4=0.25
+train_dataset_file = '/workspace/all/index/train1'
+val_dataset_file = '/workspace/all/index/val1'