Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136557162> ?p ?o ?g. }
- W3136557162 abstract "Model-agnostic meta-learning (MAML) is a popular method for few-shot learning but assumes that we have access to the meta-training set. In practice, training on the meta-training set may not always be an option due to data privacy concerns, intellectual property issues, or merely lack of computing resources. In this paper, we consider the novel problem of repurposing pretrained MAML checkpoints to solve new few-shot classification tasks. Because of the potential distribution mismatch, the original MAML steps may no longer be optimal. Therefore we propose an alternative meta-testing procedure and combine MAML gradient steps with adversarial training and uncertainty-based stepsize adaptation. Our method outperforms vanilla MAML on same-domain and cross-domains benchmarks using both SGD and Adam optimizers and shows improved robustness to the choice of base stepsize." @default.
- W3136557162 created "2021-03-29" @default.
- W3136557162 creator A5000174418 @default.
- W3136557162 creator A5027377615 @default.
- W3136557162 creator A5042730335 @default.
- W3136557162 creator A5051076050 @default.
- W3136557162 date "2021-03-16" @default.
- W3136557162 modified "2023-10-03" @default.
- W3136557162 title "Repurposing Pretrained Models for Robust Out-of-domain Few-Shot Learning" @default.
- W3136557162 cites W1522301498 @default.
- W3136557162 cites W1607019360 @default.
- W3136557162 cites W1849277567 @default.
- W3136557162 cites W2002427601 @default.
- W3136557162 cites W2099471712 @default.
- W3136557162 cites W2104094955 @default.
- W3136557162 cites W2164411961 @default.
- W3136557162 cites W2243397390 @default.
- W3136557162 cites W2533598788 @default.
- W3136557162 cites W2593768305 @default.
- W3136557162 cites W2601450892 @default.
- W3136557162 cites W2604763608 @default.
- W3136557162 cites W2753160622 @default.
- W3136557162 cites W2806143712 @default.
- W3136557162 cites W2963207607 @default.
- W3136557162 cites W2963233928 @default.
- W3136557162 cites W2963238274 @default.
- W3136557162 cites W2963341924 @default.
- W3136557162 cites W2963580001 @default.
- W3136557162 cites W2964105864 @default.
- W3136557162 cites W2964253222 @default.
- W3136557162 cites W2979920800 @default.
- W3136557162 cites W2982573148 @default.
- W3136557162 cites W2996623013 @default.
- W3136557162 cites W2996653965 @default.
- W3136557162 cites W3011267989 @default.
- W3136557162 cites W3035743198 @default.
- W3136557162 cites W3039883906 @default.
- W3136557162 cites W3043072433 @default.
- W3136557162 hasPublicationYear "2021" @default.
- W3136557162 type Work @default.
- W3136557162 sameAs 3136557162 @default.
- W3136557162 citedByCount "0" @default.
- W3136557162 crossrefType "posted-content" @default.
- W3136557162 hasAuthorship W3136557162A5000174418 @default.
- W3136557162 hasAuthorship W3136557162A5027377615 @default.
- W3136557162 hasAuthorship W3136557162A5042730335 @default.
- W3136557162 hasAuthorship W3136557162A5051076050 @default.
- W3136557162 hasConcept C104317684 @default.
- W3136557162 hasConcept C119857082 @default.
- W3136557162 hasConcept C134306372 @default.
- W3136557162 hasConcept C154945302 @default.
- W3136557162 hasConcept C162324750 @default.
- W3136557162 hasConcept C177264268 @default.
- W3136557162 hasConcept C185592680 @default.
- W3136557162 hasConcept C187736073 @default.
- W3136557162 hasConcept C18903297 @default.
- W3136557162 hasConcept C199360897 @default.
- W3136557162 hasConcept C2776434776 @default.
- W3136557162 hasConcept C2780451532 @default.
- W3136557162 hasConcept C2781002164 @default.
- W3136557162 hasConcept C33923547 @default.
- W3136557162 hasConcept C36503486 @default.
- W3136557162 hasConcept C37736160 @default.
- W3136557162 hasConcept C41008148 @default.
- W3136557162 hasConcept C51632099 @default.
- W3136557162 hasConcept C519536355 @default.
- W3136557162 hasConcept C55493867 @default.
- W3136557162 hasConcept C63479239 @default.
- W3136557162 hasConcept C774472 @default.
- W3136557162 hasConcept C86803240 @default.
- W3136557162 hasConcept C95623464 @default.
- W3136557162 hasConceptScore W3136557162C104317684 @default.
- W3136557162 hasConceptScore W3136557162C119857082 @default.
- W3136557162 hasConceptScore W3136557162C134306372 @default.
- W3136557162 hasConceptScore W3136557162C154945302 @default.
- W3136557162 hasConceptScore W3136557162C162324750 @default.
- W3136557162 hasConceptScore W3136557162C177264268 @default.
- W3136557162 hasConceptScore W3136557162C185592680 @default.
- W3136557162 hasConceptScore W3136557162C187736073 @default.
- W3136557162 hasConceptScore W3136557162C18903297 @default.
- W3136557162 hasConceptScore W3136557162C199360897 @default.
- W3136557162 hasConceptScore W3136557162C2776434776 @default.
- W3136557162 hasConceptScore W3136557162C2780451532 @default.
- W3136557162 hasConceptScore W3136557162C2781002164 @default.
- W3136557162 hasConceptScore W3136557162C33923547 @default.
- W3136557162 hasConceptScore W3136557162C36503486 @default.
- W3136557162 hasConceptScore W3136557162C37736160 @default.
- W3136557162 hasConceptScore W3136557162C41008148 @default.
- W3136557162 hasConceptScore W3136557162C51632099 @default.
- W3136557162 hasConceptScore W3136557162C519536355 @default.
- W3136557162 hasConceptScore W3136557162C55493867 @default.
- W3136557162 hasConceptScore W3136557162C63479239 @default.
- W3136557162 hasConceptScore W3136557162C774472 @default.
- W3136557162 hasConceptScore W3136557162C86803240 @default.
- W3136557162 hasConceptScore W3136557162C95623464 @default.
- W3136557162 hasOpenAccess W3136557162 @default.
- W3136557162 hasRelatedWork W2804962156 @default.
- W3136557162 hasRelatedWork W2805251476 @default.
- W3136557162 hasRelatedWork W2946605054 @default.
- W3136557162 hasRelatedWork W2974885182 @default.