Manuscript received June 20, 2025; accepted July 27, 2025; published October 15, 2025.
Abstract—This study proposes a camera control system for gesture recognition on embedded platforms using Tiny Machine Learning (TinyML). The system uses the ESP32 microcontroller and the MPU6050 inertial measurement unit (IMU) to collect gesture data, and employs a lightweight neural network model deployed with TensorFlow Lite for local processing, eliminating the reliance on the cloud. Experimental results show that the gesture recognition accuracy reached 95.2%, the wireless communication packet success rate was as high as 99.3%, and the system was optimized for power consumption during continuous operation. This work verifies the practical application capabilities of TinyML in the Internet of Things, addressing key challenges in edge computing, including computational resource limitations and real-time performance requirements, and providing a framework for developing responsive and privacy-focused control interfaces in resource-constrained environments.
Keywords—TinyML, Gesture recognition, Camera control, Embedded systems, ESP32, MPU6050, TensorFlow Lite, Real-time processing
Cite: Xiang Chenghao, "Camera Control System Based on TinyML Gesture Recognition," International Journal of Engineering and Technology, vol. 17, no. 3, pp. 183-188, 2025.
Copyright © 2025 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (
CC BY 4.0). 
