Matches in SemOpenAlex for { <https://semopenalex.org/work/W4308236280> ?p ?o ?g. }
Showing items 1 to 75 of
75
with 100 items per page.
- W4308236280 abstract "Computer vision datasets usually present long-tailed training distributions where the classes are not represented with the same number of training samples. This so-called class imbalance problem hinders the proper learning of inference models, biasing them towards over-represented classes and decreasing their generalization. Adopted solutions to tackle the effect of class imbalance are based on weighting the training loss according to the number of class samples, leading to regimes where low-represented classes guide the learning just accounting for their cardinal number. To also incorporate class complexity in the process, we propose a novel training scheme called CCL: Class-wise Curriculum Learning. Classes are first sorted based on a difficulty criterion which not only accounts for the number of training samples but also for their training outcomes. The curriculum is then used to guide the training: easy classes are fed first and—incrementally, the more difficult ones are added. The proposed approach is validated for image classification using long-tailed datasets. Results show that when the proposed Class-wise Curriculum Learning scheme is used, trained models outperform specific state-of-the-art methods devoted to handle the class imbalance problem. The code, data and reported models described along this paper are publicly available at https://github.com/vpulab/CCL" @default.
- W4308236280 created "2022-11-09" @default.
- W4308236280 creator A5084485828 @default.
- W4308236280 creator A5086848513 @default.
- W4308236280 date "2022-10-16" @default.
- W4308236280 modified "2023-10-06" @default.
- W4308236280 title "CCL: Class-Wise Curriculum Learning for Class Imbalance Problems" @default.
- W4308236280 cites W2148143831 @default.
- W4308236280 cites W2296073425 @default.
- W4308236280 cites W2440599146 @default.
- W4308236280 cites W2963351448 @default.
- W4308236280 cites W2963691377 @default.
- W4308236280 cites W3024124733 @default.
- W4308236280 cites W3034601242 @default.
- W4308236280 cites W3041497966 @default.
- W4308236280 cites W4210667047 @default.
- W4308236280 doi "https://doi.org/10.1109/icip46576.2022.9897273" @default.
- W4308236280 hasPublicationYear "2022" @default.
- W4308236280 type Work @default.
- W4308236280 citedByCount "1" @default.
- W4308236280 countsByYear W43082362802023 @default.
- W4308236280 crossrefType "proceedings-article" @default.
- W4308236280 hasAuthorship W4308236280A5084485828 @default.
- W4308236280 hasAuthorship W4308236280A5086848513 @default.
- W4308236280 hasConcept C111919701 @default.
- W4308236280 hasConcept C119857082 @default.
- W4308236280 hasConcept C126838900 @default.
- W4308236280 hasConcept C134306372 @default.
- W4308236280 hasConcept C154945302 @default.
- W4308236280 hasConcept C15744967 @default.
- W4308236280 hasConcept C177148314 @default.
- W4308236280 hasConcept C183115368 @default.
- W4308236280 hasConcept C19417346 @default.
- W4308236280 hasConcept C2776214188 @default.
- W4308236280 hasConcept C2777212361 @default.
- W4308236280 hasConcept C33923547 @default.
- W4308236280 hasConcept C41008148 @default.
- W4308236280 hasConcept C47177190 @default.
- W4308236280 hasConcept C71924100 @default.
- W4308236280 hasConcept C77618280 @default.
- W4308236280 hasConcept C98045186 @default.
- W4308236280 hasConceptScore W4308236280C111919701 @default.
- W4308236280 hasConceptScore W4308236280C119857082 @default.
- W4308236280 hasConceptScore W4308236280C126838900 @default.
- W4308236280 hasConceptScore W4308236280C134306372 @default.
- W4308236280 hasConceptScore W4308236280C154945302 @default.
- W4308236280 hasConceptScore W4308236280C15744967 @default.
- W4308236280 hasConceptScore W4308236280C177148314 @default.
- W4308236280 hasConceptScore W4308236280C183115368 @default.
- W4308236280 hasConceptScore W4308236280C19417346 @default.
- W4308236280 hasConceptScore W4308236280C2776214188 @default.
- W4308236280 hasConceptScore W4308236280C2777212361 @default.
- W4308236280 hasConceptScore W4308236280C33923547 @default.
- W4308236280 hasConceptScore W4308236280C41008148 @default.
- W4308236280 hasConceptScore W4308236280C47177190 @default.
- W4308236280 hasConceptScore W4308236280C71924100 @default.
- W4308236280 hasConceptScore W4308236280C77618280 @default.
- W4308236280 hasConceptScore W4308236280C98045186 @default.
- W4308236280 hasFunder F4320313831 @default.
- W4308236280 hasLocation W43082362801 @default.
- W4308236280 hasOpenAccess W4308236280 @default.
- W4308236280 hasPrimaryLocation W43082362801 @default.
- W4308236280 hasRelatedWork W2043505748 @default.
- W4308236280 hasRelatedWork W2885094885 @default.
- W4308236280 hasRelatedWork W2899084033 @default.
- W4308236280 hasRelatedWork W2961085424 @default.
- W4308236280 hasRelatedWork W3046775127 @default.
- W4308236280 hasRelatedWork W4285260836 @default.
- W4308236280 hasRelatedWork W4286629047 @default.
- W4308236280 hasRelatedWork W4306321456 @default.
- W4308236280 hasRelatedWork W4306674287 @default.
- W4308236280 hasRelatedWork W4224009465 @default.
- W4308236280 isParatext "false" @default.
- W4308236280 isRetracted "false" @default.
- W4308236280 workType "article" @default.