Abstract: Open-vocabulary semantic segmentation (OVSS) aims to segment an image into regions of corresponding semantic vocabularies, without being limited to a predefined set of object categories.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results