Placeholder chapter page. Sections are produced by the book-skills pipeline.
Sections
- 61.1 TinyML design constraints
- 61.2 Feature extraction on microcontrollers
- 61.3 Quantized networks and TinyNAS/MCUNet
- 61.4 Runtimes: LiteRT/TF-Lite-Micro, microTVM, ExecuTorch
- 61.5 Wake-up cascades and always-on sensing
- 61.6 Firmware integration and testing on hardware
- 61.7 Field update and rollback; MLPerf Tiny
Lab 61
deploy a motion classifier to an MCU (or a faithful embedded simulation) and benchmark it.