environment
Introduction
cartoonizer is an open source image/video cartoonization project. The author also integrated flask web framework in the project. We can easily deploy it to the network for our own or others’ use.
online experience
The official website provides a web site for online experience. The address is: https://cartoonize-lkqov62dia-de.a.run.app/ . I found a photo of the goddess. After uploading it, it will be processed in a while. Come Let’s see the effect before and after
local practice
First, create a brand new virtual environment
conda create -n cartoonizer python=3.7 conda activate cartoonizer
Then pull the code locally
git clone https://github.com/experience-ml/cartoonize.git cd cartoonize # 按照外部的依赖库pip install -r requirements.txt
After the dependencies are completed, the service can be started. This is a web service that uses flask framework
Then, open http://127.0.0.1:8080 in your browser
Upload image processing
Overall, it is the same as the online version provided by the author
It should be noted here that video processing cannot take effect, and the original code is developed based on linux , so if you want to experience video animation, you need to switch to linux operating system.
FAQ
ImportError: cannot import name ‘Markup’ from ‘jinja2’
Upgrade flask
pip install -U flask
NotImplementedError: Cannot convert a symbolic Tensor (strided_slice:0) to a numpy array.
Here you need to lower the version of numpy
pip install numpy==1.19.5
References
- https://cartoonize-lkqov62dia-de.a.run.app/
- https://github.com/experience-ml/cartoonize
- https://github.com/experience-ml/cartoonize/issues/9
This article is transferred from https://xugaoxiang.com/2023/03/17/image-video-cartoonizer/
This site is only for collection, and the copyright belongs to the original author.