Matches in SemOpenAlex for { <https://semopenalex.org/work/W4226267130> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W4226267130 abstract "Recent progress in few-shot learning promotes a more realistic cross-domain setting, where the source and target datasets are in different domains. Due to the domain gap and disjoint label spaces between source and target datasets, their shared knowledge is extremely limited. This encourages us to explore more information in the target domain rather than to overly elaborate training strategies on the source domain as in many existing methods. Hence, we start from a generic representation pre-trained by a cross-entropy loss and a conventional distance-based classifier, along with an image retrieval view, to employ a re-ranking process to calibrate a target distance matrix by discovering the k-reciprocal neighbours within the task. Assuming the pre-trained representation is biased towards the source, we construct a non-linear subspace to minimise task-irrelevant features therewithin while keep more transferrable discriminative information by a hyperbolic tangent transformation. The calibrated distance in this target-aware non-linear sub-space is complementary to that in the pre-trained representation. To impose such distance calibration information onto the pre-trained representation, a Kullback-Leibler divergence loss is employed to gradually guide the model towards the calibrated distance-based distribution. Extensive evaluations on eight target domains show that this target ranking calibration process can improve conventional distance-based classifiers in few-shot learning." @default.
- W4226267130 created "2022-05-05" @default.
- W4226267130 creator A5023834700 @default.
- W4226267130 creator A5039302902 @default.
- W4226267130 creator A5084959430 @default.
- W4226267130 creator A5086290413 @default.
- W4226267130 date "2022-06-01" @default.
- W4226267130 modified "2023-10-16" @default.
- W4226267130 title "Ranking Distance Calibration for Cross-Domain Few-Shot Learning" @default.
- W4226267130 cites W2039118116 @default.
- W4226267130 cites W2043409787 @default.
- W4226267130 cites W2076188996 @default.
- W4226267130 cites W2100398441 @default.
- W4226267130 cites W2140844045 @default.
- W4226267130 cites W2170881581 @default.
- W4226267130 cites W2242818826 @default.
- W4226267130 cites W2963322158 @default.
- W4226267130 cites W2964105864 @default.
- W4226267130 cites W3012255272 @default.
- W4226267130 cites W3035102141 @default.
- W4226267130 cites W3035143213 @default.
- W4226267130 cites W3035539956 @default.
- W4226267130 cites W3159907343 @default.
- W4226267130 cites W3173086070 @default.
- W4226267130 cites W3176793698 @default.
- W4226267130 cites W3187295906 @default.
- W4226267130 cites W3187947593 @default.
- W4226267130 cites W3189329097 @default.
- W4226267130 cites W3189656202 @default.
- W4226267130 cites W3203015511 @default.
- W4226267130 cites W3203979321 @default.
- W4226267130 cites W4214611271 @default.
- W4226267130 cites W4214764301 @default.
- W4226267130 cites W4288073030 @default.
- W4226267130 doi "https://doi.org/10.1109/cvpr52688.2022.00889" @default.
- W4226267130 hasPublicationYear "2022" @default.
- W4226267130 type Work @default.
- W4226267130 citedByCount "4" @default.
- W4226267130 countsByYear W42262671302023 @default.
- W4226267130 crossrefType "proceedings-article" @default.
- W4226267130 hasAuthorship W4226267130A5023834700 @default.
- W4226267130 hasAuthorship W4226267130A5039302902 @default.
- W4226267130 hasAuthorship W4226267130A5084959430 @default.
- W4226267130 hasAuthorship W4226267130A5086290413 @default.
- W4226267130 hasBestOaLocation W42262671302 @default.
- W4226267130 hasConcept C106301342 @default.
- W4226267130 hasConcept C114614502 @default.
- W4226267130 hasConcept C119857082 @default.
- W4226267130 hasConcept C121332964 @default.
- W4226267130 hasConcept C124101348 @default.
- W4226267130 hasConcept C153180895 @default.
- W4226267130 hasConcept C154945302 @default.
- W4226267130 hasConcept C171752962 @default.
- W4226267130 hasConcept C189430467 @default.
- W4226267130 hasConcept C2639959 @default.
- W4226267130 hasConcept C32834561 @default.
- W4226267130 hasConcept C33923547 @default.
- W4226267130 hasConcept C41008148 @default.
- W4226267130 hasConcept C45340560 @default.
- W4226267130 hasConcept C62520636 @default.
- W4226267130 hasConcept C95623464 @default.
- W4226267130 hasConcept C97931131 @default.
- W4226267130 hasConceptScore W4226267130C106301342 @default.
- W4226267130 hasConceptScore W4226267130C114614502 @default.
- W4226267130 hasConceptScore W4226267130C119857082 @default.
- W4226267130 hasConceptScore W4226267130C121332964 @default.
- W4226267130 hasConceptScore W4226267130C124101348 @default.
- W4226267130 hasConceptScore W4226267130C153180895 @default.
- W4226267130 hasConceptScore W4226267130C154945302 @default.
- W4226267130 hasConceptScore W4226267130C171752962 @default.
- W4226267130 hasConceptScore W4226267130C189430467 @default.
- W4226267130 hasConceptScore W4226267130C2639959 @default.
- W4226267130 hasConceptScore W4226267130C32834561 @default.
- W4226267130 hasConceptScore W4226267130C33923547 @default.
- W4226267130 hasConceptScore W4226267130C41008148 @default.
- W4226267130 hasConceptScore W4226267130C45340560 @default.
- W4226267130 hasConceptScore W4226267130C62520636 @default.
- W4226267130 hasConceptScore W4226267130C95623464 @default.
- W4226267130 hasConceptScore W4226267130C97931131 @default.
- W4226267130 hasFunder F4320321001 @default.
- W4226267130 hasLocation W42262671301 @default.
- W4226267130 hasLocation W42262671302 @default.
- W4226267130 hasLocation W42262671303 @default.
- W4226267130 hasOpenAccess W4226267130 @default.
- W4226267130 hasPrimaryLocation W42262671301 @default.
- W4226267130 hasRelatedWork W1967062067 @default.
- W4226267130 hasRelatedWork W2024160000 @default.
- W4226267130 hasRelatedWork W2061273563 @default.
- W4226267130 hasRelatedWork W2142361057 @default.
- W4226267130 hasRelatedWork W2285052147 @default.
- W4226267130 hasRelatedWork W2321141263 @default.
- W4226267130 hasRelatedWork W2729514902 @default.
- W4226267130 hasRelatedWork W2743258233 @default.
- W4226267130 hasRelatedWork W2773500201 @default.
- W4226267130 hasRelatedWork W4287995534 @default.
- W4226267130 isParatext "false" @default.
- W4226267130 isRetracted "false" @default.
- W4226267130 workType "article" @default.