Rethinking Learnable Tree Filter for Generic Feature Transform Lin Song 1 Y anwei Li

Neural Information Processing Systems 

The Learnable Tree Filter presents a remarkable approach to model structure-preserving relations for semantic segmentation. Nevertheless, the intrinsic geometric constraint forces it to focus on the regions with close spatial distance, hindering the effective long-range interactions.

Similar Docs  Excel Report  more

TitleSimilaritySource
None found