search

TuSimple / TuSimple-DUC

Understanding Convolution for Semantic Segmentation
https://arxiv.org/abs/1702.08502
Apache License 2.0
610 stars 118 forks source link

Where to find the train-list "train_bigger_patch.lst" #2

Closed thomasdic2000 closed 7 years ago

thomasdic2000 commented 7 years ago

I tried to use the default 'imglists/train.lst,' but seems format not recognized. Could you please provide the train-list you used? Thanks.

GrassSunFlower commented 7 years ago

You can try using data_prep/get_cityscapes_list.py to generate the data list

acgtyrant commented 6 years ago

@GrassSunFlower How to decide image_path and label_path in get_cityscapes_list.py?

As my understanding to this file, I should pass some arguments as below:

root = '/home/acgtyrant/BigDatas/Cityscapes'
image_path = 'leftImg8bit/train'
label_path = 'gtFine/train'

And cityscaples_loader.read_data will read the full image path and label path as item[0] and item[1].

However, utils.get_single_image_duc will still join item[0] and data_dir again while the item[0] is the full path already, the same for item[1].

GrassSunFlower commented 6 years ago

Yes, it is working this way. You can try change the content of data_dir into something like . or modify the output of the get_cityscapes_list output.