AdaMNet：Adaptive Compression Framework for MCU
a software-hardware co-design model deployment process based on a group of compression operators and Dynamic Voltage Frequency Scaling (DVFS) for microcontroller (MCU) devices according to dynamic context.
Model Architecture Diagram：
Description： Microcontrollers (or MCUs) have become one of the key deployment platforms for edge intelligence applications due to their small size, low cost, low power consumption, and high integration. The algorithm explores and analyzes the real hardware platform overhead and resource limitations of the microcontroller, and utilizes the “sstructublue block” concept to implement a model loading design based on compressed operator groups. It collaborates with platform-level discrete voltage and frequency configuration to build san adaptive search space for dynamic scenarios. The design at the software and hardware level is relatively independent and does not affect each other, and has high scalability.