Build and deploy your AI-powered search applications in seconds with Jina’s open-source tech stack.
1from jina import Document, Flow 2from jina.types.document.generators import from_files 3import numpy as np 4 5flow = (Flow() 6 .add(uses="jinahub+docker://ImageTorchEncoder") 7 .add(uses="jinahub+docker://SimpleIndexer")) 8 9with flow: 10 images =  11 for doc in from_files('fashion_images/*.png'): 12 doc.convert_image_uri_to_blob() 13 doc.blob = doc.blob.astype(np.uint8) 14 images.append(doc) 15 flow.index(images) 16 17 query = Document(uri='t-shirt.png') 18 query.convert_image_uri_to_blob() 19 query.blob = doc.blob.astype(np.uint8) 20 flow.search(query) 21
Search with images:
Jina lets developers and enterprises build search systems that are fast, scalable, and work with any kind of data.
"Creating search experiences to understand the intent and the contextual meaning of queries is no longer an abstract research topic. Jina AI democratises neural search and enables companies like us to bring semantic search capabilities to our clients. Right here, right now."
CEO & Co-founder, WordLift
"Jina allows us to focus on our product and tech design instead of being tangled into 3D model neural search service setup, opening up new business potential and user experience."
Co-Founder & CTO, Yahaha
About our latest releases, events, and news.