Reproduce “Talk to Me, Jarvis: An Open-Source Edge-Deployable Voice Assistant Framework for Autonomous Racecars” 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
Daniel Henel, Frederik Werner, Alexander Langmann, Johannes Betz
Jarvis maps locally transcribed speech to predefined autonomous-racing commands, using a wake-word pipeline and a QLoRA-fine-tuned, 4-bit Mistral-7B classifier. On a 17-class, partly synthetic text dataset, it reached 97.63% accuracy with 1.39 seconds average classifier inference on an RTX 3060 Laptop GPU. This does not establish end-to-end voice, safety, or race-performance gains.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.