Matches in SemOpenAlex for { <https://semopenalex.org/work/W2990135557> ?p ?o ?g. }
- W2990135557 abstract "We propose a novel non-parametric method for cross-modal retrieval which is applied on top of precomputed image and text embeddings. By combining our method with standard approaches for building image and text encoders, trained independently with a self-supervised classification objective, we create a baseline model which outperforms most existing methods on a challenging image-to-recipe task. We also use our method for comparing image and text encoders trained using different modern approaches, thus addressing the issues hindering the developments of novel methods for cross-modal recipe retrieval. We demonstrate how to use the insights from model comparison and extend our baseline model with standard triplet loss that improves SoTA on the Recipe1M dataset by a large margin, while using only precomputed features and with much less complexity than existing methods." @default.
- W2990135557 created "2019-12-05" @default.
- W2990135557 creator A5018327261 @default.
- W2990135557 creator A5053738305 @default.
- W2990135557 creator A5058754318 @default.
- W2990135557 creator A5073503574 @default.
- W2990135557 date "2019-11-28" @default.
- W2990135557 modified "2023-09-27" @default.
- W2990135557 title "Dividing and Conquering Cross-Modal Recipe Retrieval: from Nearest Neighbours Baselines to SoTA" @default.
- W2990135557 cites W12634471 @default.
- W2990135557 cites W1486649854 @default.
- W2990135557 cites W1522301498 @default.
- W2990135557 cites W1861492603 @default.
- W2990135557 cites W1964073652 @default.
- W2990135557 cites W1976591483 @default.
- W2990135557 cites W2025341678 @default.
- W2990135557 cites W204268067 @default.
- W2990135557 cites W2064675550 @default.
- W2990135557 cites W2099471712 @default.
- W2990135557 cites W2100398441 @default.
- W2990135557 cites W2108598243 @default.
- W2990135557 cites W2125143698 @default.
- W2990135557 cites W2131744502 @default.
- W2990135557 cites W2153579005 @default.
- W2990135557 cites W2184188583 @default.
- W2990135557 cites W2194775991 @default.
- W2990135557 cites W2206370378 @default.
- W2990135557 cites W2266728343 @default.
- W2990135557 cites W2493916176 @default.
- W2990135557 cites W2530422462 @default.
- W2990135557 cites W2531634031 @default.
- W2990135557 cites W2565041183 @default.
- W2990135557 cites W2583892095 @default.
- W2990135557 cites W2605649771 @default.
- W2990135557 cites W2613010453 @default.
- W2990135557 cites W2737041163 @default.
- W2990135557 cites W2747515317 @default.
- W2990135557 cites W2765440071 @default.
- W2990135557 cites W2765707332 @default.
- W2990135557 cites W2784874046 @default.
- W2990135557 cites W2897152025 @default.
- W2990135557 cites W2948037078 @default.
- W2990135557 cites W2953328958 @default.
- W2990135557 cites W2960416371 @default.
- W2990135557 cites W2962964995 @default.
- W2990135557 cites W2963055199 @default.
- W2990135557 cites W2963341956 @default.
- W2990135557 cites W2963499246 @default.
- W2990135557 cites W2963804993 @default.
- W2990135557 cites W2963997278 @default.
- W2990135557 cites W2964216321 @default.
- W2990135557 cites W2970464537 @default.
- W2990135557 cites W2982573303 @default.
- W2990135557 cites W2994818707 @default.
- W2990135557 cites W2997283289 @default.
- W2990135557 cites W3003735286 @default.
- W2990135557 cites W3008072721 @default.
- W2990135557 cites W3017133364 @default.
- W2990135557 cites W3035032757 @default.
- W2990135557 cites W3105114834 @default.
- W2990135557 cites W3117993946 @default.
- W2990135557 cites W3130985505 @default.
- W2990135557 cites W343636949 @default.
- W2990135557 cites W639708223 @default.
- W2990135557 hasPublicationYear "2019" @default.
- W2990135557 type Work @default.
- W2990135557 sameAs 2990135557 @default.
- W2990135557 citedByCount "3" @default.
- W2990135557 countsByYear W29901355572020 @default.
- W2990135557 countsByYear W29901355572021 @default.
- W2990135557 crossrefType "posted-content" @default.
- W2990135557 hasAuthorship W2990135557A5018327261 @default.
- W2990135557 hasAuthorship W2990135557A5053738305 @default.
- W2990135557 hasAuthorship W2990135557A5058754318 @default.
- W2990135557 hasAuthorship W2990135557A5073503574 @default.
- W2990135557 hasConcept C105795698 @default.
- W2990135557 hasConcept C111368507 @default.
- W2990135557 hasConcept C111919701 @default.
- W2990135557 hasConcept C115961682 @default.
- W2990135557 hasConcept C117251300 @default.
- W2990135557 hasConcept C118505674 @default.
- W2990135557 hasConcept C119857082 @default.
- W2990135557 hasConcept C12725497 @default.
- W2990135557 hasConcept C127313418 @default.
- W2990135557 hasConcept C153180895 @default.
- W2990135557 hasConcept C154945302 @default.
- W2990135557 hasConcept C166957645 @default.
- W2990135557 hasConcept C185592680 @default.
- W2990135557 hasConcept C188027245 @default.
- W2990135557 hasConcept C205649164 @default.
- W2990135557 hasConcept C2778671685 @default.
- W2990135557 hasConcept C33923547 @default.
- W2990135557 hasConcept C41008148 @default.
- W2990135557 hasConcept C71139939 @default.
- W2990135557 hasConcept C774472 @default.
- W2990135557 hasConceptScore W2990135557C105795698 @default.
- W2990135557 hasConceptScore W2990135557C111368507 @default.
- W2990135557 hasConceptScore W2990135557C111919701 @default.
- W2990135557 hasConceptScore W2990135557C115961682 @default.
- W2990135557 hasConceptScore W2990135557C117251300 @default.