Reproduce “Luganda Speech Intent Recognition for IoT Applications” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Andrew Katumba, Sudi Murindanyi, John Trevor Kasule, Elvis Mugume
The authors build a locally processed Luganda voice-command recognizer that connects speech to household devices. The system extracts MFCCs and uses a CNN to classify 20 intents, sending commands over MQTT through a Raspberry Pi to ESP32 controllers. With 13 MFCCs, augmentation, and global max pooling, it reports 87.36% accuracy, a prototype result rather than evidence of everyday reliability.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.