We introduce cost aggregation to open-vocabulary semantic segmentation, which jointly aggregates both image and text modalities within the matching cost. For further details and visualization results, ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results