Aidhwi

Aidhwi - AI Driven Hand Writing Input.

Purpose

An attempt to recreate a handwriting recognition input method from scratch in C++.

Components

• Front-end
• Character recognition neural network
• TODO

Dependencies

Front-end:

• SDL2 — Input and display abstraction
• GLAD — GL loader
• Dear ImGui — Immediate mode GUI library
• GLM — GL math

Building

Clone the repository and follow the instructions for your platform:

Note: This repository has several dependencies as Git submodules. They should be initialized by running git submodule init && git submodule update before proceeding. Alternatively you can pass --recurse-submodules to git clone command when cloning the repo.

Note: This project uses CMake to generate build scripts. Make sure you have CMake version 3.13 or higher.

Note: The following instructions assume that you are using Ninja as a build system. Get Ninja if you want to follow the instructions exactly.

Desktop

1. Ensure you have all required dependencies installed and configured (specifically SDL).
2. Create a build directory (preferably somewhere outside of the project structure) and cd to it.
3. Run cmake -G Ninja "/path/to/this/repo" -B ..
4. Run ninja to build the generated project.

To test the build copy the res directory (and all required DLLs if on Windows) to your build and run ./aidhwi.

Web

1. Get Emscripten and its dependencies: emscripten.org
2. Optional: go through the tutorial to check if it is set up properly
3. Create a build directory (preferably somewhere outside of the project structure) and cd to it.
4. Run cmake -G Ninja "/path/to/this/repo" "-DCMAKE_TOOLCHAIN_FILE=/path/to/Emscripten.cmake" -B ..
   Note: Emscripten.cmake can usually be found in /path/to/emsdk/upstream/emscripten/cmake/Modules/Platform/.
5. Run ninja to build the generated project.

To test the build start any http server in the build directory to serve on port 8000 (or any other available one) and navigate to http://localhost:8000/aidhwi.html.