Semantic Segmentation Example

Annotation

bash

labelme data_annotated --labels labels.txt --validate-label exact --config '{shape_color: {mode: auto, auto: {shift: -2}}}'

Convert to VOC-format Dataset

bash

# It generates:
#   - data_dataset_voc/JPEGImages
#   - data_dataset_voc/SegmentationClass
#   - data_dataset_voc/SegmentationClassNpy
#   - data_dataset_voc/SegmentationClassVisualization
./labelme2voc.py data_annotated data_dataset_voc --labels labels.txt --noobject

Fig 1. JPEG image (left), PNG label (center), JPEG label visualization (right)

Note that the label file contains only very low label values (ex. 0, 4, 14), and 255 indicates the __ignore__ label value (-1 in the npy file).
You can see the label PNG file by following.

bash

../tutorial/draw_label_png.py data_dataset_voc/SegmentationClass/2011_000003.png