AI Paper Atlas

Research map / Image generation / Image restoration

HyperTransformer: A Textural and Spectral Feature Fusion Transformer for Pansharpening

Wele Gedara Chaminda Bandara, Vishal M. Patel

CVPR 2022 · Image generation › Image restoration · approach: Pansharpening and spectral fusion · 193 citations (OpenAlex)

Check your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use cross-feature multi-head soft-attention instead of bypassing MHFSA (ablation)HyperTransformer · Pavia CenterCC: 0.981 -> 0.989Improved
Use 1 global descriptor instead of 16 (ablation)HyperTransformer · Pavia CenterCC: 0.989 -> 0.975Hurt
Use 2 global descriptors instead of 16 (ablation)HyperTransformer · Pavia CenterCC: 0.989 -> 0.976Hurt
Use 8 global descriptors instead of 16 (ablation)HyperTransformer · Pavia CenterCC: 0.989 -> 0.987Hurt
Use 32 global descriptors instead of 16 (ablation)HyperTransformer · Pavia CenterCC: 0.989 -> 0.988Hurt
Use 64 global descriptors instead of 16 (ablation)HyperTransformer · Pavia CenterCC: 0.989 -> 0.987Hurt
Apply HyperTransformer at all three scales rather than at ×4 only (ablation)HyperTransformer · Pavia CenterCC: 0.986 -> 0.989Improved
Add ×1 to the ×4 scale configuration (ablation)HyperTransformer · Pavia CenterCC: 0.986 -> 0.988Improved
Add ×2 to the ×4 scale configuration (ablation)HyperTransformer · Pavia CenterCC: 0.986 -> 0.988Improved
Add VGG perceptual loss to L1 training (ablation)HyperTransformer · Pavia CenterCC: 0.987 -> 0.988Improved
Add transfer perceptual loss to L1 and VGG perceptual loss (ablation)HyperTransformer · Pavia CenterCC: 0.988 -> 0.989Improved
Replace the strongest baseline with HyperTransformerPansharpening methods · Pavia CenterCC: 0.980 -> 0.989Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy