Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
You can not select more than 25 topics
Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
ZhiHao Cao
e5d8c2a6d5
|
2 years ago | |
|---|---|---|
| imsp | 3 years ago | |
| .gitignore | 3 years ago | |
| LICENSE | 3 years ago | |
| README.md | 2 years ago | |
| fruit.jpg | 3 years ago | |
| requirements.txt | 3 years ago | |
| road.jpg | 3 years ago | |
| setup.py | 3 years ago | |
README.md
Paddle-IMSP
An Image Search and Pair system base on PaddlePaddle.
Search Web Image by text and pair similar images by image.
Base project：【Paddle-CLIP】.
Project Information
-
This project is limited to non-commercial scenarios
-
All the retrieved and matched web images are from the website: 【unsplash】
-
If you need to use the original image, please move to 【unsplash】 Download
-
Thanks to 【unsplash】 open source large-scale image dataset
-
Thanks to 【openai/CLIP】 open source model code and pretrained model parameters
Install Package
- Install by pip：
$ pip install paddleimsp
- Install by wheel package：【Releases Packages】
Quick Start
from imsp import IMSP
# Load the engine
# The first load will automatically download the pretrained model and images database
imsp_engine = IMSP()
# Search Web Image by text
# You can use Chinese or English
photo_urls = imsp_engine.im_search('sky')
Preview photos:
Original image：Unsplash Link
Original image：Unsplash Link
Original image：Unsplash Link
# Pair similar images by image
photo_urls = imsp_engine.im_pair('fruit.jpg')
Preview photos:
Original image：Unsplash Link
Original image：Unsplash Link
Original image：Unsplash Link
# Use the images and texts to search and pair images
photo_urls = imsp_engine.im_search_pair(images=['road.jpg'], texts=['cars', 'peoples'], topk=5)
Preview photos:
Original image：Unsplash Link
Original image：Unsplash Link
Original image：Unsplash Link
Contact us
Email : agentmaker@163.com
QQ Group : 1005109853
No Description
Python Text
