Matches in SemOpenAlex for { <https://semopenalex.org/work/W2965729909> ?p ?o ?g. }
- W2965729909 abstract "At Pinterest, we utilize image embeddings throughout our search and recommendation systems to help our users navigate through visual content by powering experiences like browsing of related content and searching for exact products for shopping. In this work we describe a multi-task deep metric learning system to learn a single unified image embedding which can be used to power our multiple visual search products. The solution we present not only allows us to train for multiple application objectives in a single deep neural network architecture, but takes advantage of correlated information in the combination of all training data from each application to generate a unified embedding that outperforms all specialized embeddings previously deployed for each product. We discuss the challenges of handling images from different domains such as camera photos, high quality web images, and clean product catalog images. We also detail how to jointly train for multiple product objectives and how to leverage both engagement data and human labeled data. In addition, our trained embeddings can also be binarized for efficient storage and retrieval without compromising precision and recall. Through comprehensive evaluations on offline metrics, user studies, and online A/B experiments, we demonstrate that our proposed unified embedding improves both relevance and engagement of our visual search products for both browsing and searching purposes when compared to existing specialized embeddings. Finally, the deployment of the unified embedding at Pinterest has drastically reduced the operation and engineering cost of maintaining multiple embeddings while improving quality." @default.
- W2965729909 created "2019-08-13" @default.
- W2965729909 creator A5025289132 @default.
- W2965729909 creator A5027383005 @default.
- W2965729909 creator A5057865121 @default.
- W2965729909 creator A5063812292 @default.
- W2965729909 creator A5068607680 @default.
- W2965729909 date "2019-08-05" @default.
- W2965729909 modified "2023-09-28" @default.
- W2965729909 title "Learning a Unified Embedding for Visual Search at Pinterest" @default.
- W2965729909 cites W1686810756 @default.
- W2965729909 cites W1839408879 @default.
- W2965729909 cites W2021354639 @default.
- W2965729909 cites W2090334843 @default.
- W2965729909 cites W2100142570 @default.
- W2965729909 cites W2108598243 @default.
- W2965729909 cites W2121339428 @default.
- W2965729909 cites W2157364932 @default.
- W2965729909 cites W2163605009 @default.
- W2965729909 cites W2194775991 @default.
- W2965729909 cites W2339391301 @default.
- W2965729909 cites W2549139847 @default.
- W2965729909 cites W2555897561 @default.
- W2965729909 cites W2593507512 @default.
- W2965729909 cites W2618011341 @default.
- W2965729909 cites W2622263826 @default.
- W2965729909 cites W2625758617 @default.
- W2965729909 cites W2689134854 @default.
- W2965729909 cites W2752782242 @default.
- W2965729909 cites W2767434619 @default.
- W2965729909 cites W2768591600 @default.
- W2965729909 cites W2786649424 @default.
- W2965729909 cites W2795783309 @default.
- W2965729909 cites W2798512429 @default.
- W2965729909 cites W2798930779 @default.
- W2965729909 cites W2903034054 @default.
- W2965729909 cites W2950855501 @default.
- W2965729909 cites W2952373945 @default.
- W2965729909 cites W2963026686 @default.
- W2965729909 cites W2963258075 @default.
- W2965729909 cites W2965158072 @default.
- W2965729909 cites W3099206234 @default.
- W2965729909 hasPublicationYear "2019" @default.
- W2965729909 type Work @default.
- W2965729909 sameAs 2965729909 @default.
- W2965729909 citedByCount "0" @default.
- W2965729909 crossrefType "posted-content" @default.
- W2965729909 hasAuthorship W2965729909A5025289132 @default.
- W2965729909 hasAuthorship W2965729909A5027383005 @default.
- W2965729909 hasAuthorship W2965729909A5057865121 @default.
- W2965729909 hasAuthorship W2965729909A5063812292 @default.
- W2965729909 hasAuthorship W2965729909A5068607680 @default.
- W2965729909 hasConcept C108583219 @default.
- W2965729909 hasConcept C119857082 @default.
- W2965729909 hasConcept C136764020 @default.
- W2965729909 hasConcept C153083717 @default.
- W2965729909 hasConcept C154945302 @default.
- W2965729909 hasConcept C158154518 @default.
- W2965729909 hasConcept C162324750 @default.
- W2965729909 hasConcept C176217482 @default.
- W2965729909 hasConcept C17744445 @default.
- W2965729909 hasConcept C199539241 @default.
- W2965729909 hasConcept C21547014 @default.
- W2965729909 hasConcept C23123220 @default.
- W2965729909 hasConcept C41008148 @default.
- W2965729909 hasConcept C41608201 @default.
- W2965729909 hasConcept C71901391 @default.
- W2965729909 hasConceptScore W2965729909C108583219 @default.
- W2965729909 hasConceptScore W2965729909C119857082 @default.
- W2965729909 hasConceptScore W2965729909C136764020 @default.
- W2965729909 hasConceptScore W2965729909C153083717 @default.
- W2965729909 hasConceptScore W2965729909C154945302 @default.
- W2965729909 hasConceptScore W2965729909C158154518 @default.
- W2965729909 hasConceptScore W2965729909C162324750 @default.
- W2965729909 hasConceptScore W2965729909C176217482 @default.
- W2965729909 hasConceptScore W2965729909C17744445 @default.
- W2965729909 hasConceptScore W2965729909C199539241 @default.
- W2965729909 hasConceptScore W2965729909C21547014 @default.
- W2965729909 hasConceptScore W2965729909C23123220 @default.
- W2965729909 hasConceptScore W2965729909C41008148 @default.
- W2965729909 hasConceptScore W2965729909C41608201 @default.
- W2965729909 hasConceptScore W2965729909C71901391 @default.
- W2965729909 hasOpenAccess W2965729909 @default.
- W2965729909 hasRelatedWork W1968594412 @default.
- W2965729909 hasRelatedWork W1981068360 @default.
- W2965729909 hasRelatedWork W2007510844 @default.
- W2965729909 hasRelatedWork W2330065172 @default.
- W2965729909 hasRelatedWork W2608152660 @default.
- W2965729909 hasRelatedWork W2769547064 @default.
- W2965729909 hasRelatedWork W2792276797 @default.
- W2965729909 hasRelatedWork W2909206463 @default.
- W2965729909 hasRelatedWork W2911086437 @default.
- W2965729909 hasRelatedWork W2950398970 @default.
- W2965729909 hasRelatedWork W2950500537 @default.
- W2965729909 hasRelatedWork W2950628404 @default.
- W2965729909 hasRelatedWork W2965433310 @default.
- W2965729909 hasRelatedWork W3043363188 @default.
- W2965729909 hasRelatedWork W3080750010 @default.
- W2965729909 hasRelatedWork W3127952366 @default.
- W2965729909 hasRelatedWork W3128537296 @default.