Bootstrapped Language Image Pretraining

GPTKB entity

Statements (40)
Predicate Object
gptkbp:instanceOf gptkb:model
gptkbp:abbreviation gptkb:BLIP
gptkbp:approach contrastive learning
language modeling
image-text matching
gptkbp:architecture gptkb:Vision_Transformer
Transformer decoder
Transformer encoder
gptkbp:arXivID 2201.12086
gptkbp:author gptkb:Silvio_Savarese
Dongxu Li
Junnan Li
Steven Hoi
gptkbp:developedBy gptkb:Salesforce_Research
gptkbp:field computer vision
natural language processing
multimodal learning
gptkbp:firstPublished 2022
https://www.w3.org/2000/01/rdf-schema#label Bootstrapped Language Image Pretraining
gptkbp:input gptkb:illustrator
gptkb:text
gptkbp:license gptkb:MIT_License
gptkbp:memiliki_tugas image captioning
image-text retrieval
visual question answering
gptkbp:notablePublication gptkb:BLIP:_Bootstrapped_Language-Image_Pre-training_for_Unified_Vision-Language_Understanding_and_Generation
gptkbp:output text descriptions
image-text representations
retrieval results
gptkbp:pretrainingMethod bootstrapped pretraining
gptkbp:relatedTo gptkb:ALIGN
gptkb:CLIP
gptkb:ViLT
gptkbp:repository https://github.com/salesforce/BLIP
gptkbp:trainer gptkb:COCO
gptkb:Flickr30k
gptkb:VQA
NoCaps
gptkbp:bfsParent gptkb:BLIP
gptkbp:bfsLayer 7