FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387492750> ?p ?o ?g. }

• W4387492750 endingPage "121526" @default.
• W4387492750 startingPage "121526" @default.
• W4387492750 abstract "Recently multimodal medical pretraining models play a significant role in automatic medical image and text analysis that has wide social and economical impact in healthcare. Despite being able to be quickly transferred to downstream tasks, the models are greatly limited due to the fact that these models can only be pretrained with professional medical image-text datasets, which usually contain a very small number of samples. In this work We propose MITER (Medical Image-Text Joint adaptive Pretraining), a joint adaptive pretraining framework via multi-level contrastive learning to overcome this limitation by pretraining image and text models for medical domain and utilizing existing models pretrained on generic data, which contain enormous number of samples. MITER features two types of objectives to solve the problem. The first type is uni-modal objectives that pretrain the models with medical images and text separately on uni-modal tasks. The other type is a cross-modal objective that pretrains jointly, allowing the models to influence each other on cross-modal tasks. We also introduce a strategy to dynamically select hard negative samples during the training process for better performance. Experimental results over four medical tasks, image-report retrieval, multi-label image classification, visual question answering, and report generation, show that our MITER framework solves the limitation problem by greatly outperforming existing benchmark models on all the tasks. The source code of our framework is available online.2" @default.
• W4387492750 created "2023-10-11" @default.
• W4387492750 creator A5000122982 @default.
• W4387492750 creator A5036242008 @default.
• W4387492750 creator A5036257681 @default.
• W4387492750 creator A5044548403 @default.
• W4387492750 creator A5048112222 @default.
• W4387492750 creator A5052441498 @default.
• W4387492750 creator A5067305024 @default.
• W4387492750 creator A5091189673 @default.
• W4387492750 date "2023-10-01" @default.
• W4387492750 modified "2023-10-12" @default.
• W4387492750 title "MITER: Medical Image-TExt joint adaptive pretRaining with multi-level contrastive learning" @default.
• W4387492750 cites W1956340063 @default.
• W4387492750 cites W2101105183 @default.
• W4387492750 cites W2152772232 @default.
• W4387492750 cites W2396881363 @default.
• W4387492750 cites W2561981131 @default.
• W4387492750 cites W2901466771 @default.
• W4387492750 cites W2912664121 @default.
• W4387492750 cites W2963466845 @default.
• W4387492750 cites W2963967185 @default.
• W4387492750 cites W3013277995 @default.
• W4387492750 cites W3022046290 @default.
• W4387492750 cites W3030790048 @default.
• W4387492750 cites W3094950914 @default.
• W4387492750 cites W3101156210 @default.
• W4387492750 cites W3138559567 @default.
• W4387492750 cites W3164654615 @default.
• W4387492750 cites W3173220247 @default.
• W4387492750 cites W3181252431 @default.
• W4387492750 cites W3186890061 @default.
• W4387492750 cites W3196437584 @default.
• W4387492750 cites W3198570286 @default.
• W4387492750 cites W3201906559 @default.
• W4387492750 cites W3213233983 @default.
• W4387492750 cites W4220685350 @default.
• W4387492750 cites W4225404354 @default.
• W4387492750 cites W4290716922 @default.
• W4387492750 cites W4376865522 @default.
• W4387492750 doi "https://doi.org/10.1016/j.eswa.2023.121526" @default.
• W4387492750 hasPublicationYear "2023" @default.
• W4387492750 type Work @default.
• W4387492750 citedByCount "0" @default.
• W4387492750 crossrefType "journal-article" @default.
• W4387492750 hasAuthorship W4387492750A5000122982 @default.
• W4387492750 hasAuthorship W4387492750A5036242008 @default.
• W4387492750 hasAuthorship W4387492750A5036257681 @default.
• W4387492750 hasAuthorship W4387492750A5044548403 @default.
• W4387492750 hasAuthorship W4387492750A5048112222 @default.
• W4387492750 hasAuthorship W4387492750A5052441498 @default.
• W4387492750 hasAuthorship W4387492750A5067305024 @default.
• W4387492750 hasAuthorship W4387492750A5091189673 @default.
• W4387492750 hasConcept C111919701 @default.
• W4387492750 hasConcept C115961682 @default.
• W4387492750 hasConcept C119857082 @default.
• W4387492750 hasConcept C127413603 @default.
• W4387492750 hasConcept C13280743 @default.
• W4387492750 hasConcept C134306372 @default.
• W4387492750 hasConcept C153180895 @default.
• W4387492750 hasConcept C154945302 @default.
• W4387492750 hasConcept C170154142 @default.
• W4387492750 hasConcept C18555067 @default.
• W4387492750 hasConcept C185592680 @default.
• W4387492750 hasConcept C185798385 @default.
• W4387492750 hasConcept C188027245 @default.
• W4387492750 hasConcept C205649164 @default.
• W4387492750 hasConcept C33923547 @default.
• W4387492750 hasConcept C36503486 @default.
• W4387492750 hasConcept C41008148 @default.
• W4387492750 hasConcept C71139939 @default.
• W4387492750 hasConcept C98045186 @default.
• W4387492750 hasConceptScore W4387492750C111919701 @default.
• W4387492750 hasConceptScore W4387492750C115961682 @default.
• W4387492750 hasConceptScore W4387492750C119857082 @default.
• W4387492750 hasConceptScore W4387492750C127413603 @default.
• W4387492750 hasConceptScore W4387492750C13280743 @default.
• W4387492750 hasConceptScore W4387492750C134306372 @default.
• W4387492750 hasConceptScore W4387492750C153180895 @default.
• W4387492750 hasConceptScore W4387492750C154945302 @default.
• W4387492750 hasConceptScore W4387492750C170154142 @default.
• W4387492750 hasConceptScore W4387492750C18555067 @default.
• W4387492750 hasConceptScore W4387492750C185592680 @default.
• W4387492750 hasConceptScore W4387492750C185798385 @default.
• W4387492750 hasConceptScore W4387492750C188027245 @default.
• W4387492750 hasConceptScore W4387492750C205649164 @default.
• W4387492750 hasConceptScore W4387492750C33923547 @default.
• W4387492750 hasConceptScore W4387492750C36503486 @default.
• W4387492750 hasConceptScore W4387492750C41008148 @default.
• W4387492750 hasConceptScore W4387492750C71139939 @default.
• W4387492750 hasConceptScore W4387492750C98045186 @default.
• W4387492750 hasLocation W43874927501 @default.
• W4387492750 hasOpenAccess W4387492750 @default.
• W4387492750 hasPrimaryLocation W43874927501 @default.
• W4387492750 hasRelatedWork W2028665553 @default.
• W4387492750 hasRelatedWork W2086519370 @default.
• W4387492750 hasRelatedWork W2087343574 @default.
• W4387492750 hasRelatedWork W2130974462 @default.

• next