This is a Kotlin app that uses a machine learning model hosted on a server to translate rough sketches into realistic portraits. The app allows users to create a sketch by drawing on the screen, and then sends a screenshot of the drawing to the server for processing. The server converts the sketch into a portrait and sends it back to the app, which displays the portrait to the user.
- Android device with a touchscreen and internet connectivity
The app uses Image-to-image translation with conditional adversarial networks hosted on a server to translate the sketch into a portrait. The model was trained on a large dataset of sketches and portraits, and is able to generate realistic portraits from rough sketches.
The app communicates with the server using HTTP requests and JSON payloads. It sends a screenshot of the sketch to the server, and the server returns the resulting portrait in the form of a base64 String.
The app was developed in Kotlin and makes use of various Android libraries and frameworks, such as the Android SDK, and Retrofit.
