Morse Decoder is a machine learning application that interprets morse code for people with disabilities. The project is proudly being developed for AceCenter, a UK charity that specializes in Augmentative and Alternative Communication and Assistive Technology.
The main outcome for this project was to be able to take an input-agnostic morse code signal, validate that it is a proper voluntary command, and interpret output not only limited to characters, but also including “control patterns” that allow it to interact with digital devices in the form of keys such as Delete, Enter, Tab, etc. Eventually, we aim to pair it with optional modules for fast typing words and error correction. For this, we are using Docker for containerization, FastAPI for handling module requests, Natural Language Processing modules for fast-typing experiments, and Python for most BackEnd programming.