Resource-Constrained Perception: Deploying Quantized Deep Learning Models on Microcontrollers for Real-Time Robotic Vision
DOI:
https://doi.org/10.62411/jimat.v3i2.17734Keywords:
Quantization, microcontrollers, Edge AI, Robotic Vision, Lightweight CNNAbstract
Robotic vision enables machines to "see" and interpret the world around them. It can help with tasks such as obstacle avoidance, navigation, and object detection. This kind of work is well suited to modern deep learning but requires a lot of memory, power, and processing speed. Most of the small robots do not have a powerful computer. However, they have a low-cost microcontroller. This is a smaller computer. These devices are very resource-constrained. It becomes difficult to run regular deep learning models. This research work proposes to address the above issue with model quantization. When a model number is not a nice integer. It is quantized, i.e., rounded to a simpler integer with 8 bits. Such a reduction reduces memory consumption and increases the computing speed. This work is a comparison of two popular approaches. They are Post-Training Quantization (PTQ) and Quantization-Aware Training (QAT). Researchers construct a small and simple ConvNet that has only 28,069 parameters. It is examined on two popular microcontrollers: ARM Cortex M7 and ESP32. Both methods have been found to be effective. By reducing the use of the flash memory by approximately 71% and RAM by 60%, quantization cuts down memory usage. It doubles the processing speed and reduces power consumption by 35 to 40 percent. QAT retains 99% of the original model accuracy. The advantages of PTQ are that it is faster to set up, but with slightly less accuracy. The model performs well in real time on both platforms, with no assistance from the clouds. This work shows deep learning can run well on low-power devices. It helps make smart, low-cost robots more widely available.References
Chen, S., Li, Y., & Kwok, N. M. (2011). Active vision in robotic systems: A survey of recent developments. The International Journal of Robotics Research, 30(11), 1343-1377.
Pérez, L., Rodríguez, Í., Rodríguez, N., Usamentiaga, R., & García, D. F. (2016). Robot guidance using machine vision techniques in industrial environments: A comparative review. Sensors, 16(3), 335.
Manakitsa, N., Maraslidis, G. S., Moysis, L., & Fragulis, G. F. (2024). A review of machine learning and deep learning for object detection, semantic segmentation, and human action recognition in machine and robotic vision. Technologies, 12(2), 15.
Menghani, G. (2023). Efficient deep learning: A survey on making deep learning models smaller, faster, and better. ACM Computing Surveys, 55(12), 1-37.
Sun, Y. et al. (2021). Evaluating performance, power and energy of deep neural networks on CPUs and GPUs. Communications in Computer and Information Science, 1494, Springer.
Sharma, A., Jain, A., Gupta, P., & Chowdary, V. (2020). Machine learning applications for precision agriculture: A comprehensive review. IEEE Access, 9, 4843-4873.
Gupte, S. (2025). Evaluating microcontrollers for embedded and IoT applications: Criteria, trade-offs, and selection framework. International Journal for Research in Applied Science and Engineering Technology, 13(7), 655-660.
Immonen, R., & Hämäläinen, T. (2022). Tiny machine learning for resource-constrained microcontrollers. Journal of Sensors, 2022, 1-11.
Li et al. (2024). Contemporary advances in neural network quantization: A survey. International Joint Conference on Neural Networks (IJCNN), Yokohama, Japan.
Liu D, Zhu Y, Liu Z, Liu Y, Han C, Tian J, Li R & Yi W (2025). A survey of model compression techniques: Past, present, and future. Frontiers in Robotics and AI, 12:1518965.
Capogrosso, L., Cunico, F., Cheng, D.S., Fummi, F. & Cristani, F. (2024). A machine learning-oriented survey on tiny machine learning. IEEE Access, 12, 23406-23426.
Chen, L., Li, G., Xie, W., Tan, J., Li, Y., Pu, J., Chen, L., Gan, D., & Shi, W. (2024). A survey of computer vision detection, visual SLAM algorithms, and their applications in energy-efficient autonomous systems. Energies, 17(20), 5177.
Barbosa, J. P. de A. et al. (2015). ROS, Android and cloud robotics: How to make a powerful low cost robot. International Conference on Advanced Robotics (ICAR), Istanbul, Turkey.
Hossain, M.E., Tarafder, M.T.R., Ahmed, N., Noman, A.A., Sarkar, M.I. & Hossain, Z. (2023). Integrating AI with edge computing and cloud services for real-time data processing and decision making. International Journal of Multidisciplinary Sciences and Arts, 2(4), 252-261.
Schizas, N., Karras, A., Karras, C., & Sioutas, S. (2022). TinyML for ultra-low power AI and large scale IoT deployments: A systematic review. Future Internet, 14(12), 363.
Maciá-Lillo, A., Barrachina, S., Fabregat, G., & Dolz, M. F. (2024). Optimizing convolutions for deep learning inference on ARM Cortex-M processors. IEEE Internet of Things Journal, 11(15), 26203-26219.
Cheng, H., Zhang, M. & Shi, J.Q. (2024). A survey on deep neural network pruning: Taxonomy, comparison, analysis, and recommendations. IEEE Transactions on Pattern Analysis and Machine Intelligence, 46, 10558-10578.
Novac, P.-E., Boukli Hacene, G., Pegatoquet, A., Miramond, B., & Gripon, V. (2021). Quantization and deployment of deep neural networks on microcontrollers. Sensors, 21(9), 2984.
Saleem, T. J., Hardan, S., Shaaban, M. A., & Yaqub, M. (2026). Deep learning for healthcare: Trends in lightweight and efficient solutions. SSRN: https://ssrn.com/abstract=6100966.
Xiao, G., Lin, J., Seznec, M., Wu, H., Demouth, J. & Han, S. (2023). SmoothQuant: Accurate and efficient post-training quantization for large language models. Proceedings of Machine Learning Research, 202, 38087-38099.
Saha, S. S., Sandha, S. S., & Srivastava, M. (2022). Machine learning for microcontroller-class hardware: A review. IEEE Sensors Journal, 22(22), 21362-21390.
Shuvo, M. M. H., Islam, S. K., Cheng, J., & Morshed, B. I. (2022). Efficient acceleration of deep learning inference on resource-constrained edge devices: A review. Proceedings of the IEEE, 111(1), 42-91.
Downloads
Published
Issue
Section
License
Authors who publish their articles in this journal agree to the following conditions:
- Copyright remains with the author and gives the JIMAT journal the right as first priority to publish the article under a Creative Commons Attribution License which allows articles to be shared with acknowledgment of the author of the article and this journal as the place of publication.
- Authors can distribute their published articles non-exclusively (for example: in university repositories or in books) with notification or acknowledgment of publication in JIMAT.
- Authors are permitted to list their work online (for example: on a personal website or in a university repository) before and after the submission process (see The Effect of Open Access).
