Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313172370> ?p ?o ?g. }
- W4313172370 endingPage "577" @default.
- W4313172370 startingPage "558" @default.
- W4313172370 abstract "Large Vision & Language models pretrained on web-scale data provide representations that are invaluable for numerous V &L problems. However, it is unclear how they can be extended to reason about user-specific visual concepts in unstructured language. This problem arises in multiple domains, from personalized image retrieval to personalized interaction with smart devices. We introduce a new learning setup called Personalized Vision & Language (PerVL) with two new benchmark datasets for retrieving and segmenting user-specific (“personalized”) concepts “in the wild”. In PerVL, one should learn personalized concepts (1) independently of the downstream task (2) allowing a pretrained model to reason about them with free language, and (3) without providing personalized negative examples. We propose an architecture for solving PerVL that operates by expanding the input vocabulary of a pretrained model with new word embeddings for the personalized concepts. The model can then simply employ them as part of a sentence. We demonstrate that our approach learns personalized visual concepts from a few examples and effectively applies them in image retrieval and semantic segmentation using rich textual queries. For example the model improves MRR by 51.1% (28.4% vs 18.8%) compared to the strongest baseline. The code and benchmark are available on github under NVlabs/PALAVRA ( https://github.com/NVlabs/PALAVRA ) and NVlabs/PerVLBenchmark ( https://github.com/NVlabs/PerVLBenchmark )." @default.
- W4313172370 created "2023-01-06" @default.
- W4313172370 creator A5017999137 @default.
- W4313172370 creator A5045719865 @default.
- W4313172370 creator A5055304028 @default.
- W4313172370 creator A5057080001 @default.
- W4313172370 creator A5080909499 @default.
- W4313172370 date "2022-01-01" @default.
- W4313172370 modified "2023-09-25" @default.
- W4313172370 title "“This Is My Unicorn, Fluffy”: Personalizing Frozen Vision-Language Representations" @default.
- W4313172370 cites W1861492603 @default.
- W4313172370 cites W1966109131 @default.
- W4313172370 cites W1987549167 @default.
- W4313172370 cites W2008835805 @default.
- W4313172370 cites W2029385256 @default.
- W4313172370 cites W2171061940 @default.
- W4313172370 cites W2173180041 @default.
- W4313172370 cites W2596142952 @default.
- W4313172370 cites W2765301947 @default.
- W4313172370 cites W2795151422 @default.
- W4313172370 cites W2797733588 @default.
- W4313172370 cites W2889986507 @default.
- W4313172370 cites W2923366293 @default.
- W4313172370 cites W2962689421 @default.
- W4313172370 cites W2963088515 @default.
- W4313172370 cites W2964105864 @default.
- W4313172370 cites W2966070748 @default.
- W4313172370 cites W2967819436 @default.
- W4313172370 cites W2972897806 @default.
- W4313172370 cites W2982260276 @default.
- W4313172370 cites W3026458074 @default.
- W4313172370 cites W3034332915 @default.
- W4313172370 cites W3035342403 @default.
- W4313172370 cites W3081143589 @default.
- W4313172370 cites W3101226978 @default.
- W4313172370 cites W3119510203 @default.
- W4313172370 cites W3132926949 @default.
- W4313172370 cites W3143107425 @default.
- W4313172370 cites W3175270254 @default.
- W4313172370 cites W3176196997 @default.
- W4313172370 cites W3176484337 @default.
- W4313172370 cites W3198675127 @default.
- W4313172370 cites W3201762919 @default.
- W4313172370 cites W3204532479 @default.
- W4313172370 cites W3205249428 @default.
- W4313172370 cites W3212108239 @default.
- W4313172370 cites W4214700015 @default.
- W4313172370 cites W4214926101 @default.
- W4313172370 cites W4226091687 @default.
- W4313172370 cites W4286611278 @default.
- W4313172370 cites W4288083516 @default.
- W4313172370 doi "https://doi.org/10.1007/978-3-031-20044-1_32" @default.
- W4313172370 hasPublicationYear "2022" @default.
- W4313172370 type Work @default.
- W4313172370 citedByCount "2" @default.
- W4313172370 countsByYear W43131723702023 @default.
- W4313172370 crossrefType "book-chapter" @default.
- W4313172370 hasAuthorship W4313172370A5017999137 @default.
- W4313172370 hasAuthorship W4313172370A5045719865 @default.
- W4313172370 hasAuthorship W4313172370A5055304028 @default.
- W4313172370 hasAuthorship W4313172370A5057080001 @default.
- W4313172370 hasAuthorship W4313172370A5080909499 @default.
- W4313172370 hasBestOaLocation W43131723702 @default.
- W4313172370 hasConcept C13280743 @default.
- W4313172370 hasConcept C137293760 @default.
- W4313172370 hasConcept C138885662 @default.
- W4313172370 hasConcept C154945302 @default.
- W4313172370 hasConcept C177264268 @default.
- W4313172370 hasConcept C184337299 @default.
- W4313172370 hasConcept C185798385 @default.
- W4313172370 hasConcept C18903297 @default.
- W4313172370 hasConcept C199360897 @default.
- W4313172370 hasConcept C204321447 @default.
- W4313172370 hasConcept C205649164 @default.
- W4313172370 hasConcept C23123220 @default.
- W4313172370 hasConcept C2776760102 @default.
- W4313172370 hasConcept C2777530160 @default.
- W4313172370 hasConcept C2777601683 @default.
- W4313172370 hasConcept C2780734062 @default.
- W4313172370 hasConcept C41008148 @default.
- W4313172370 hasConcept C41895202 @default.
- W4313172370 hasConcept C44291984 @default.
- W4313172370 hasConcept C86803240 @default.
- W4313172370 hasConcept C89600930 @default.
- W4313172370 hasConcept C90805587 @default.
- W4313172370 hasConceptScore W4313172370C13280743 @default.
- W4313172370 hasConceptScore W4313172370C137293760 @default.
- W4313172370 hasConceptScore W4313172370C138885662 @default.
- W4313172370 hasConceptScore W4313172370C154945302 @default.
- W4313172370 hasConceptScore W4313172370C177264268 @default.
- W4313172370 hasConceptScore W4313172370C184337299 @default.
- W4313172370 hasConceptScore W4313172370C185798385 @default.
- W4313172370 hasConceptScore W4313172370C18903297 @default.
- W4313172370 hasConceptScore W4313172370C199360897 @default.
- W4313172370 hasConceptScore W4313172370C204321447 @default.
- W4313172370 hasConceptScore W4313172370C205649164 @default.
- W4313172370 hasConceptScore W4313172370C23123220 @default.
- W4313172370 hasConceptScore W4313172370C2776760102 @default.