-
Large-Margin Representation Learning for Texture Classification
The proposed approach uses a short sequence of convolutional layers (CLs) to learn representation for texture classification. -
SPair-71k: A large-scale benchmark for semantic correspondence
The proposed method, dubbed Dynamic Hyperpixel Flow, learns to compose hypercolumn features on the fly by selecting a small number of relevant layers from a deep convolutional... -
Learning to Compose Hypercolumns for Visual Correspondence
Feature representation plays a crucial role in visual correspondence, and recent methods for image matching resort to deeply stacked convolutional layers. -
HSViT: Horizontally Scalable Vision Transformer
This paper introduces a horizontally scalable vision transformer (HSViT) scheme with a novel image-level feature embedding. The design of HSViT preserves the inductive bias from...