Review Article
Tiny Machine Learning for Resource-Constrained Microcontrollers
Table 1
TinyML application features in IoT devices.
| Feature | Increased | Decreased |
| Energy usage | CPU | Radio, overall | Processor | Throughput | Latency | Memory | Allocation complexity | Footprint | Data | Security, privacy | Transfer need | ML model | Optimization demand | Size, accuracy | Costs | Software dev. | Hardware |
|
|