Jina: Neural Search Engine For Images, Videos, Audios

Jina is a neural search engine that enables users to efficiently search for images, videos, and audios. It leverages popular libraries like openclip, haystack, towhee, and Milvus to provide pre-trained models and workflows through Jina Hub. Additionally, VCED offers tutorials in machine learning topics.

---

openclip

haystack

tutorial: build QA pipeline with no dependencies with haystack

towhee

milvus

visit jina hub to get multiple embedding models and workflows

jina import video/image/text

finetuner: text to image search via clip

datawhale provides tutorials on machine learning, also provide book materials, topics are: numpy, matplotlib, pandas,

vced: holy gift from datawhale able to edit video by text, video auto editor, cutter

VCED 可以通过你的文字描述来自动识别视频中相符合的片段进行视频剪辑。该项目基于跨模态搜索与向量检索技术搭建，通过前后端分离的模式，帮助你快速的接触新一代搜索技术。

jina:

https://github.com/jina-ai/jina/

documentation:

https://docs.jina.ai

quick demos:

dress Fashion image search: jina hello fashion

robot QA chatbot: pip install “jina[demo]” && jina hello chatbot

newspaper Multimodal search: pip install “jina[demo]” && jina hello multimodal

fork_and_knife Fork the source of a demo to your folder: jina hello fork fashion ../my-proj/

Create a new Jina project: jina new hello-jina

ai video metadata generation: