Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385570100> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4385570100 abstract "Multimodal machine learning is a cutting-edge field that explores ways to incorporate information from multiple sources into models. As more multimodal data becomes available, this field has become increasingly relevant. This work focuses on two key challenges in multimodal machine learning. The first is finding efficient ways to combine information from different data types. The second is that often, one modality (e.g., text) is stronger and more relevant, making it difficult to identify meaningful patterns in the weaker modality (e.g., image). Our approach focuses on more effectively exploiting the weaker modality while dynamically regularizing the loss function. First, we introduce a new two-stream model called Multimodal BERT-ViT, which features a novel intra-CLS token fusion. Second, we utilize a dynamic adjustment that maintains a balance between specialization and generalization during the training to avoid overfitting, which we devised. We add this dynamic adjustment to the Unsupervised Data Augmentation (UDA) framework. We evaluate the effectiveness of these proposals on the task of multi-label movie genre classification using the Moviescope and MM-IMDb datasets. The evaluation revealed that our proposal offers substantial benefits, while simultaneously enabling us to harness the weaker modality without compromising the information provided by the stronger." @default.
- W4385570100 created "2023-08-05" @default.
- W4385570100 creator A5054006751 @default.
- W4385570100 creator A5055252239 @default.
- W4385570100 creator A5092596152 @default.
- W4385570100 date "2023-01-01" @default.
- W4385570100 modified "2023-09-25" @default.
- W4385570100 title "Dynamic Regularization in UDA for Transformers in Multimodal Classification" @default.
- W4385570100 doi "https://doi.org/10.18653/v1/2023.acl-long.485" @default.
- W4385570100 hasPublicationYear "2023" @default.
- W4385570100 type Work @default.
- W4385570100 citedByCount "0" @default.
- W4385570100 crossrefType "proceedings-article" @default.
- W4385570100 hasAuthorship W4385570100A5054006751 @default.
- W4385570100 hasAuthorship W4385570100A5055252239 @default.
- W4385570100 hasAuthorship W4385570100A5092596152 @default.
- W4385570100 hasBestOaLocation W43855701001 @default.
- W4385570100 hasConcept C119857082 @default.
- W4385570100 hasConcept C154945302 @default.
- W4385570100 hasConcept C202444582 @default.
- W4385570100 hasConcept C22019652 @default.
- W4385570100 hasConcept C26517878 @default.
- W4385570100 hasConcept C2776135515 @default.
- W4385570100 hasConcept C2780226545 @default.
- W4385570100 hasConcept C33923547 @default.
- W4385570100 hasConcept C38652104 @default.
- W4385570100 hasConcept C41008148 @default.
- W4385570100 hasConcept C48145219 @default.
- W4385570100 hasConcept C50644808 @default.
- W4385570100 hasConcept C9652623 @default.
- W4385570100 hasConceptScore W4385570100C119857082 @default.
- W4385570100 hasConceptScore W4385570100C154945302 @default.
- W4385570100 hasConceptScore W4385570100C202444582 @default.
- W4385570100 hasConceptScore W4385570100C22019652 @default.
- W4385570100 hasConceptScore W4385570100C26517878 @default.
- W4385570100 hasConceptScore W4385570100C2776135515 @default.
- W4385570100 hasConceptScore W4385570100C2780226545 @default.
- W4385570100 hasConceptScore W4385570100C33923547 @default.
- W4385570100 hasConceptScore W4385570100C38652104 @default.
- W4385570100 hasConceptScore W4385570100C41008148 @default.
- W4385570100 hasConceptScore W4385570100C48145219 @default.
- W4385570100 hasConceptScore W4385570100C50644808 @default.
- W4385570100 hasConceptScore W4385570100C9652623 @default.
- W4385570100 hasLocation W43855701001 @default.
- W4385570100 hasOpenAccess W4385570100 @default.
- W4385570100 hasPrimaryLocation W43855701001 @default.
- W4385570100 hasRelatedWork W1996541855 @default.
- W4385570100 hasRelatedWork W2940336242 @default.
- W4385570100 hasRelatedWork W2953328427 @default.
- W4385570100 hasRelatedWork W2989932438 @default.
- W4385570100 hasRelatedWork W3099765033 @default.
- W4385570100 hasRelatedWork W4206962509 @default.
- W4385570100 hasRelatedWork W4210794429 @default.
- W4385570100 hasRelatedWork W4313159793 @default.
- W4385570100 hasRelatedWork W4361732492 @default.
- W4385570100 hasRelatedWork W4362499066 @default.
- W4385570100 isParatext "false" @default.
- W4385570100 isRetracted "false" @default.
- W4385570100 workType "article" @default.