InstructMixup: Instruction-Guided Salient Patch Editing for Robust Data Augmentation
In image and video technologies, data augmentation is widely used to improve the generalization of deep visual models, and mixup-based strategies that interpolate between samples have become the dominant approach. However, computing informative mixing regions adds substantial overhead, and blending content across different images frequently disrupts the semantic integrity of the resulting sample. We propose \our{}, a data augmentation method that constructs challenging yet label-consistent training samples entirely within a single visual sample. \our{} first extracts multi-scale salient patche
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- FuzzyOverlapping authors or contributors · 62%keras-team/keras →
“Shared author/contributor keys: jin”
- FuzzyOverlapping authors or contributors · 62%HKUDS/LightRAG →
“Shared author/contributor keys: jin”
- LinkedLinked via arxiv author · 85%Khawar Islam →
“InstructMixup: Instruction-Guided Salient Patch Editing for Robust Data Augmentation”
- LinkedLinked via arxiv author · 85%Arif Mahmood →
“InstructMixup: Instruction-Guided Salient Patch Editing for Robust Data Augmentation”
- LinkedLinked via arxiv author · 85%Xin Jin →
“InstructMixup: Instruction-Guided Salient Patch Editing for Robust Data Augmentation”
- LinkedLinked via arxiv author · 85%Naveed Akhtar →
“InstructMixup: Instruction-Guided Salient Patch Editing for Robust Data Augmentation”