Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper re-frames texture interpolation by replacing fixed positional feature blending with appearance-aware texton matching, arguing that arbitrary spatial correspondences hinder interpolation quality for unrelated textures. By treating textures as collections of local appearance vectors (textons) and matching them based on similarity rather than location, the authors demonstrate consistent improvements in the distributional quality of interpolated midpoints across various texture pairs. The study reveals that the choice of which local appearances are mixed is as critical as the mixing strength itself, with soft matching yielding superior midpoint statistics and one-to-one assignments offering more balanced transitions.
Mixing the right local appearances, not just blending pixels, unlocks significantly higher quality texture interpolations.
Texture interpolation asks what should lie between two textures. Standard methods typically blend pixels or features at corresponding spatial locations, implicitly assuming that co-located elements should be mixed. For unrelated textures, however, this correspondence is arbitrary: a crack at one location has no particular relationship to a fiber at the same location in another image. We study whether this hidden pairing choice is itself a major determinant of interpolation quality. We represent textures as collections of local appearance vectors, or textons, and replace positional pairing with appearance-aware matching while keeping the encoder, representation, and decoder fixed. This isolates the effect of correspondence from changes in model capacity or training. Across texture pairs, appearance-aware matching consistently improves the distributional quality of interpolated midpoints over positional feature blending. Soft matching produces the strongest midpoint statistics, whereas one-to-one assignments yield more balanced transitions and closer agreement with the target endpoint. The same intervention also improves a second pretrained texture interpolator without retraining. These results identify correspondence as a first-class design choice in texture interpolation: meaningful transitions depend not only on how strongly two textures are mixed, but also on which local appearances are mixed with one another.