Data augmentation

[Ethanwl]

The data augmentation in code is not the same as the paper. There is no data augmentation in the code, but the code just use the same image four times.

[praypratyay]

Ya you are right. In image.py, the image is taken as input and just resized and returned. The code for the cropping of 9 patches exists but the condition for cropping is never satisfied. So it returns the img resized using cv2.

[zhengmiao1]

we can set train parameter as False in train.py for data augomentation

[CongWang-hnu]

我也发现了这个问题，我想问的是在论文中提到的车辆数据集是不是不能按照文章中提到的方法进行剪裁，毕竟那个数据集在使用是就是被剪裁的，再次剪裁可能会剪裁不到东西。