SPA unlocks patch-level features in CLIP for class-incremental learning via semantic-guided selection and optimal transport alignment with class descriptions, plus projectors and pseudo-feature replay to reduce forgetting.
R-dfcil: Relation-guided representation learning for data-free class incremental learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Unlocking Patch-Level Features for CLIP-Based Class-Incremental Learning
SPA unlocks patch-level features in CLIP for class-incremental learning via semantic-guided selection and optimal transport alignment with class descriptions, plus projectors and pseudo-feature replay to reduce forgetting.