Skip to content

oyzh888/quickfill

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

QuickFill

QuickFill is an AI-powered form-filling application that simplifies the process of filling out forms by using image uploads. It leverages GPT-4 Vision and AWS OCR technology to extract information from provided images and fill the forms accurately.

Currently we only support English due to the AWS TextExtractor only support En.

demo

Installation

To get started with QuickFill, you need to set up the application on your local environment. Follow these steps:

cd quickfill
pip install -r requirements.txt

API Key

Before using QuickFill, you need to configure your OpenAI and AWS credentials. These keys are essential for the application to function properly.

  1. OpenAI API Key: Set up your OpenAI API key to enable GPT-4 Vision capabilities.
  2. AWS Access Key: Configure your AWS access key to use AWS OCR services. Eg: Make sure you can access AWS TextExtraction Service. You might want to use the following command to setup AWS env
pip install boto3
pip install awscli
aws configure

Feel free to check the https://chat.openai.com/share/588c9fbd-c783-4276-9db3-43da8e4288de or search "How to setup AWS" at AWS official website.

Quickstart

To launch QuickFill, navigate to the application's directory and run the provided script. Then, access the application through your web browser:

# Start the backend
sh run.sh

# Start the frontend
cd quickfill/ui
python3 -m http.server

# Open the demo
## This one use the GPT4-Vision
http://localhost:8000

## This one uses the AWS OCR service
http://localhost:8000/ocr_fill.html

Frontend

The frontend code for QuickFill is located in the following directory:

quickfill/frontend/index.html

You can modify the frontend as per your needs.

Functionality

QuickFill mainly functions through two image uploads:

  • Image A: Upload an image containing the information from the user.
  • Image B: Upload an image of the target form that needs to be filled.

Technology Stack

  • GPT-4 Vision: Utilized for advanced image processing and understanding.
  • AWS OCR: Employed for Optical Character Recognition to accurately extract text from images.

Limitations

Currently, QuickFill only supports English, due to the reliance on AWS OCR services. This limitation is necessary to ensure accurate location and filling of the forms.

Citation

If you use QuickFill in your research or project, please cite the following paper:

@article{quickfill2024,
  title={Quickfill: AI form filling tech},
  author={Zhihao Ouyang, Haowen Guo},
  journal={github},
  year={2024}
}

Thank you for choosing QuickFill!

About

AI makes form filling easy

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published