Matches in SemOpenAlex for { <https://semopenalex.org/work/W3208601917> ?p ?o ?g. }
- W3208601917 endingPage "27" @default.
- W3208601917 startingPage "1" @default.
- W3208601917 abstract "This article introduces a two-phase deep feature engineering framework for efficient learning of semantics enhanced joint embedding, which clearly separates the deep feature engineering in data preprocessing from training the text-image joint embedding model. We use the Recipe1M dataset for the technical description and empirical validation. In preprocessing, we perform deep feature engineering by combining deep feature engineering with semantic context features derived from raw text-image input data. We leverage LSTM to identify key terms, deep NLP models from the BERT family, TextRank, or TF-IDF to produce ranking scores for key terms before generating the vector representation for each key term by using Word2vec. We leverage Wide ResNet50 and Word2vec to extract and encode the image category semantics of food images to help semantic alignment of the learned recipe and image embeddings in the joint latent space. In joint embedding learning, we perform deep feature engineering by optimizing the batch-hard triplet loss function with soft-margin and double negative sampling, taking into account also the category-based alignment loss and discriminator-based alignment loss. Extensive experiments demonstrate that our SEJE approach with deep feature engineering significantly outperforms the state-of-the-art approaches." @default.
- W3208601917 created "2021-11-08" @default.
- W3208601917 creator A5017360203 @default.
- W3208601917 creator A5037897744 @default.
- W3208601917 creator A5038310950 @default.
- W3208601917 creator A5060093535 @default.
- W3208601917 creator A5087405013 @default.
- W3208601917 date "2021-12-01" @default.
- W3208601917 modified "2023-09-26" @default.
- W3208601917 title "Learning Text-image Joint Embedding for Efficient Cross-modal Retrieval with Deep Feature Engineering" @default.
- W3208601917 cites W12634471 @default.
- W3208601917 cites W1949478088 @default.
- W3208601917 cites W1964073652 @default.
- W3208601917 cites W1978394996 @default.
- W3208601917 cites W1982635469 @default.
- W3208601917 cites W2092361219 @default.
- W3208601917 cites W2106277773 @default.
- W3208601917 cites W2137918516 @default.
- W3208601917 cites W2194775991 @default.
- W3208601917 cites W2264685884 @default.
- W3208601917 cites W2293499654 @default.
- W3208601917 cites W2526198870 @default.
- W3208601917 cites W2549139847 @default.
- W3208601917 cites W2565041183 @default.
- W3208601917 cites W2728515412 @default.
- W3208601917 cites W2737041163 @default.
- W3208601917 cites W2765440071 @default.
- W3208601917 cites W2767221212 @default.
- W3208601917 cites W2897152025 @default.
- W3208601917 cites W2948037078 @default.
- W3208601917 cites W2963055199 @default.
- W3208601917 cites W2963997278 @default.
- W3208601917 cites W3034303366 @default.
- W3208601917 cites W3035032757 @default.
- W3208601917 cites W4249992252 @default.
- W3208601917 doi "https://doi.org/10.1145/3490519" @default.
- W3208601917 hasPublicationYear "2021" @default.
- W3208601917 type Work @default.
- W3208601917 sameAs 3208601917 @default.
- W3208601917 citedByCount "1" @default.
- W3208601917 countsByYear W32086019172023 @default.
- W3208601917 crossrefType "journal-article" @default.
- W3208601917 hasAuthorship W3208601917A5017360203 @default.
- W3208601917 hasAuthorship W3208601917A5037897744 @default.
- W3208601917 hasAuthorship W3208601917A5038310950 @default.
- W3208601917 hasAuthorship W3208601917A5060093535 @default.
- W3208601917 hasAuthorship W3208601917A5087405013 @default.
- W3208601917 hasBestOaLocation W32086019172 @default.
- W3208601917 hasConcept C108583219 @default.
- W3208601917 hasConcept C119857082 @default.
- W3208601917 hasConcept C138885662 @default.
- W3208601917 hasConcept C153083717 @default.
- W3208601917 hasConcept C153180895 @default.
- W3208601917 hasConcept C154945302 @default.
- W3208601917 hasConcept C2776401178 @default.
- W3208601917 hasConcept C2776461190 @default.
- W3208601917 hasConcept C2777462759 @default.
- W3208601917 hasConcept C2778827112 @default.
- W3208601917 hasConcept C41008148 @default.
- W3208601917 hasConcept C41608201 @default.
- W3208601917 hasConcept C41895202 @default.
- W3208601917 hasConcept C52622490 @default.
- W3208601917 hasConcept C59404180 @default.
- W3208601917 hasConcept C83665646 @default.
- W3208601917 hasConceptScore W3208601917C108583219 @default.
- W3208601917 hasConceptScore W3208601917C119857082 @default.
- W3208601917 hasConceptScore W3208601917C138885662 @default.
- W3208601917 hasConceptScore W3208601917C153083717 @default.
- W3208601917 hasConceptScore W3208601917C153180895 @default.
- W3208601917 hasConceptScore W3208601917C154945302 @default.
- W3208601917 hasConceptScore W3208601917C2776401178 @default.
- W3208601917 hasConceptScore W3208601917C2776461190 @default.
- W3208601917 hasConceptScore W3208601917C2777462759 @default.
- W3208601917 hasConceptScore W3208601917C2778827112 @default.
- W3208601917 hasConceptScore W3208601917C41008148 @default.
- W3208601917 hasConceptScore W3208601917C41608201 @default.
- W3208601917 hasConceptScore W3208601917C41895202 @default.
- W3208601917 hasConceptScore W3208601917C52622490 @default.
- W3208601917 hasConceptScore W3208601917C59404180 @default.
- W3208601917 hasConceptScore W3208601917C83665646 @default.
- W3208601917 hasFunder F4320306076 @default.
- W3208601917 hasFunder F4320322725 @default.
- W3208601917 hasFunder F4320324638 @default.
- W3208601917 hasIssue "4" @default.
- W3208601917 hasLocation W32086019171 @default.
- W3208601917 hasLocation W32086019172 @default.
- W3208601917 hasOpenAccess W3208601917 @default.
- W3208601917 hasPrimaryLocation W32086019171 @default.
- W3208601917 hasRelatedWork W2546942002 @default.
- W3208601917 hasRelatedWork W2772780115 @default.
- W3208601917 hasRelatedWork W2773120646 @default.
- W3208601917 hasRelatedWork W2963749793 @default.
- W3208601917 hasRelatedWork W2967506468 @default.
- W3208601917 hasRelatedWork W2970435854 @default.
- W3208601917 hasRelatedWork W3197541072 @default.
- W3208601917 hasRelatedWork W4315777907 @default.
- W3208601917 hasRelatedWork W4315783664 @default.
- W3208601917 hasRelatedWork W4321488668 @default.