Multimodal Relation Extraction

Choose or input text, text format is as follow: {Text}*{Head Entity}*{Tail Entity}

(This system can extract information with the pre-defined schema such as /org/loc/locate_at, /per/per/peer)

Text
Image

It is limited to upload pictures in JPG, PNG and JPEG formats not exceeding 10M.

Objects

It is limited to upload at most 3 objects.

(You are recommended to use a visual grounding toolkit to extract visual objects from the origin image based on entities and entity types. For more details, please see readme on our GitHub.)

Visitors' number：
Copyright© 2022-2023 OpenKG & ZJUKG All Rights Reserved