An intra-class patch swap augmentation plus instance-to-instance KL distillation lets a single network train itself and beat several teacher-based and self-distillation baselines on image tasks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Intra-class Patch Swap for Self-Distillation
An intra-class patch swap augmentation plus instance-to-instance KL distillation lets a single network train itself and beat several teacher-based and self-distillation baselines on image tasks.