Fig. 19 | Scientific Reports

Fig. 19

From: Non-invasive blood glucose monitoring using PPG signals with various deep learning models and implementation using TinyML

Fig. 19

Workflow of embedded systems development: This diagram illustrates the process of inferencing on an embedded device, beginning with data collection and processing. The workflow involves utilizing a remote server (Secure Shell (SSH) is employed for secure communication between the local computer and the remote server) with a 3090 GPU for training and validating models. The selected model undergoes optimization through pruning and quantization, followed by compilation into a binary format using C++ compilers, preparing it for execution on the STM32 MCU. The model binary is then transferred from the internal memory of the STM32H743IIT6 to the external flash memory (W25Q256), enabling direct execution using the ’Execute in Place’ (XIP) function.

Back to article page