- Multiple camera support
- Add .clang-format to project
- Add .clang-tidy to project
- Implement IPC
- Separate program to client-server for faster recognition
- Somehow tweak existing dlib algorithm for better recognition OR
- Find better face recognition algorithm OR
- Actually use depth info from IR camera for better face recognition
- Integrate with exist desktop environments
- Add language pack support
- Replace the command line arguments with more reasonable ones.
- Somehow use https://github.com/umlaeute/v4l2loopback