LADA adds lightweight label-specific memory vectors to a frozen CLIP image encoder, removing the need for task-parameter selection and reporting state-of-the-art X-TAIL benchmark results.
A continual learning survey: Defying forgetting in classification tasks
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
-
LADA: Scalable Label-Specific CLIP Adapter for Continual Learning
LADA adds lightweight label-specific memory vectors to a frozen CLIP image encoder, removing the need for task-parameter selection and reporting state-of-the-art X-TAIL benchmark results.