Machine learning inference deployed on resource-constrained microcontrollers with kilobyte-scale RAM, milliwatt power budgets, and MHz-range processors, enabling always-on intelligent sensing in IoT devices, wearables, and embedded sensors without cloud connectivity. Requires aggressive model optimisation through INT8 quantisation, pruning, and knowledge distillation.
Semantic Classification
Content
TinyML (Machine Learning on Microcontrollers) (AI-0432) — content pending enrichment.