forked from paramoecium/road-segmentation
-
Notifications
You must be signed in to change notification settings - Fork 0
/
setup.sh
46 lines (37 loc) · 1.21 KB
/
setup.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
#!/usr/bin/env bash
# Extract
mkdir -p data/test_set
for i in {1..50}
do
cp data/test_set_images/test_$i/*.png data/test_set/
done
rm -rf data/test_set_images
# Create data directories
mkdir -p data/test_set/downsampled
mkdir -p data/training/groundtruth/downsampled
mkdir -p data/training/images/downsampled
# Creates log directories for checkpoints
mkdir -p logs
# Create output directories
mkdir -p results/CNN_Output/test/raw
mkdir -p results/CNN_Output/test/high_res_raw
mkdir -p results/CNN_Output/training/raw
mkdir -p results/CNN_Output/training/high_res_raw
mkdir -p results/CNN_Output_Baseline/test/raw
mkdir -p results/CNN_Output_Baseline/training/raw
mkdir -p results/Autoencoder_Output/test
mkdir -p results/Autoencoder_Output/train
mkdir -p logs/Autoencoder
mkdir -p results/CNN_Autoencoder_Output/test
mkdir -p results/CNN_Autoencoder_Output/train
mkdir -p logs/CNN_Autoencoder
# Create tmp directories to hold TensorFlow results
mkdir -p tmp
mkdir -p src/baseline/tmp
# Downsample images
cp data/test_set/*.png data/test_set/downsampled
cp data/training/groundtruth/*.png data/training/groundtruth/downsampled
cp data/training/images/*.png data/training/images/downsampled
cd src
python3 cilutil/resizing.py
cd ..