/facegram

[GDSC Solution Challenge] Facegram All-Part Merged Repository

Primary LanguageJupyter Notebook

Facegram

KeyWord Mongtage Generator

Our application Facegram can generate a montage to show a sketch


Project Goals

Number 11, 16 from 17 GOALS

  • Goal 16

    Facegram is made to solve a Residential Envrionment Crime.
    It can create a safer residential environment in relation to urban crime.

  • Goal 17

    Facegram can rebuild existing criminal image Databases and create Efficient Montage.
    It is expected to help solving missing cases and suspect prediction.


Main Features

스크린샷 2023-04-04 오후 3 18 50

1. Generate Face Montage With Keywords.

  • Facegram generate a new montage with several keywords.
  • Parameter: age, gender, face, hair, eyebrows, eyes, nose, mouth

스크린샷 2023-04-04 오후 4 07 48

2. Viewing History.

  • View what users have already picked.

Scenario

This Solution can be used

  1. Creating Montage based on Witness Testimony / Explanation
  2. Rebuild existing Crimninal Image Databases and create efficient Montage
  3. Public purpose like Face Recognition, Face Assume, Arrest Criminal, Search for Missing Persons etc..

Used Tech Stacks

  • Frontend: React Native, Expo
  • Backend: Kotlin, Spring, MySql, OAuth2, Gcp Storage, Gcp Compute Engine, Docker
  • A.I: Google Cloud, Pytorch, Colab, ML

How to run Code

  • Frontend: Be sure to install node js and react native settings before starting. npm start will run the application.
  • Backend: First, you have to run Mysql database using dockerfile inside source code.
    After then you can run Spring Framework code using some IDEs like IntelliJ.
  • A.I: Run the notebook file named "test.ipynb" in the order of its cells, one after the other.

Presetation Video

Video Link