Performance Evaluation of Image Classification Models on Resource-Constrained STM32 Microcontrollers

Deploying deep learning on microcontrollers offers real-time intelligence at the edge, but tight memory and compute budgets complicate design choices. This study evaluates image classification on the STM32H747I-DISCO using a compact convolutional neural network trained on five board classes (Arduino Uno, Node MCU, ESP8266-01, Micro: bit V2.0, ESP32-CAM). A small, augmented dataset (50–100 images per class) was used with standard transformations; models were quantised to int8 and deployed via STM32CubeIDE and the STM32-AI CLI. The analysis examines how input resolution (1080p vs 480p) interacts with accuracy, memory footprint, latency, and power. Four classes achieve ≥95% accuracy across both resolutions, while ESP8266-01 improves from 65.7% (1080p) to 92.3% (480p), suggesting that downsampling can suppress distracting fine-grained artefacts. Activation-buffer tuning and post-training quantisation reduce RAM from ~761 kB to ~610 kB and Flash from ~1.42 MB to ~1.20 MB without accuracy loss; 480p further lowers latency by up to 35% and power by ~20%. The findings provide a resolution-aware benchmark and practical guidance for balancing fidelity and efficiency on STM32-class MCUs, and they motivate future work with larger benchmarks, cross-platform comparisons, and pruning/distillation pipelines

STM32H747I-DISCO; Tiny ML; Edge AI; Image

1. Andrade, E., Pietrantuono, R., Machida, F., & Cotroneo, D. (2023). A Comparative Analysis of Software Aging in Image Classifiers on Cloud and Edge. IEEE Transactions on Dependable and Secure Computing, 20(1), 563–573. https://doi.org/10.1109/TDSC.2021.3139201 [Google Scholar] [Crossref]

2. Berta, R., Dabbous, A., Lazzaroni, L., Pau, D. Pietro, & Bellotti, F. (2024). Developing a TinyML Image Classifier in an Hour. IEEE Open Journal of the Industrial Electronics Society, 5, 946–960. https://doi.org/10.1109/OJIES.2024.3451959 [Google Scholar] [Crossref]

3. Chepkov, A. O., Klimachev, V. S., Korchagin, A. I., & Vlasov, A. I. (2021). Analysis of the features of image processing using the Hamming network on the STM-32 microcontroller. Journal of Physics: Conference Series, 1889(2), 022046. https://doi.org/10.1088/1742-6596/1889/2/022046 [Google Scholar] [Crossref]

4. de Vita, F., Nocera, G., Bruneo, D., Tomaselli, V., Giacalone, D., & Das, S. K. (2020). Quantitative Analysis of Deep Leaf: a Plant Disease Detector on the Smart Edge. 2020 IEEE International Conference on Smart Computing (SMARTCOMP), 49–56. https://doi.org/10.1109/SMARTCOMP50058.2020.00027 [Google Scholar] [Crossref]

5. Dominguez-Morales, J. P., Duran-Lopez, L., Gutierrez-Galan, D., Rios-Navarro, A., Linares-Barranco, A., & Jimenez-Fernandez, A. (2021). Wildlife Monitoring on the Edge: A Performance Evaluation of Embedded Neural Networks on Microcontrollers for Animal Behavior Classification. Sensors, 21(9), 2975. https://doi.org/10.3390/s21092975 [Google Scholar] [Crossref]

6. Gao, H., Zhong, S., Dong, L., & Yuan, T. (2023). Performance Analysis of Machine Learning Algorithm on GD32 Microcontrollers. 2023 6th International Conference on Artificial Intelligence and Big Data (ICAIBD), 763–767. https://doi.org/10.1109/ICAIBD57115.2023.10206295 [Google Scholar] [Crossref]

7. Han, F., Li, L., Wang, K., Feng, F., Pan, H., & Yu, D. (2016). An improved FFT architecture optimized for reconfigurable application specified processor. Proceedings - 2015 IEEE 11th International Conference on ASIC, ASICON 2015, 2, 5–8. https://doi.org/10.1109/ASICON.2015.7517201 [Google Scholar] [Crossref]

8. Howard, A., Sandler, M., Chen, B., Wang, W., Chen, L. C., Tan, M., Chu, G., Vasudevan, V., Zhu, Y., Pang, R., Le, Q., & Adam, H. (2019). Searching for mobileNetV3. Proceedings of the IEEE International Conference on Computer Vision. https://doi.org/10.1109/ICCV.2019.00140 [Google Scholar] [Crossref]

9. Lee, D., Kim, J.-S., & Hong, S. (2024). Dual-Core-Based Microcontrollers Inference Design and Performance Analysis. IEEE Access, 12, 120326–120336. https://doi.org/10.1109/ACCESS.2024.3443406 [Google Scholar] [Crossref]

10. Ma, N., Zhang, X., Zheng, H. T., & Sun, J. (2018). Shufflenet V2: Practical guidelines for efficient cnn architecture design. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 11218 LNCS. https://doi.org/10.1007/978-3-030-01264-9_8 [Google Scholar] [Crossref]

11. Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L. C. (2018). MobileNetV2: Inverted Residuals and Linear Bottlenecks. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. https://doi.org/10.1109/CVPR.2018.00474 [Google Scholar] [Crossref]

12. Svoboda, F., Fernandez-Marques, J., Liberis, E., & Lane, N. D. (2022). Deep learning on microcontrollers: A study on deployment costs and challenges. EuroMLSys 2022 - Proceedings of the 2nd European Workshop on Machine Learning and Systems, 54–63. https://doi.org/10.1145/3517207.3526978 [Google Scholar] [Crossref]

13. Thang, H. (2021). Implementation of Deep Learning Neural Network LeNet5 on STM32 Microcontroller for Image Recognition. TNU Journal of Science and Technology, 226(11), 191–199. https://doi.org/10.34238/tnu-jst.4497 [Google Scholar] [Crossref]

14. Xie, Y.-L., Lin, X.-R., & Lin, C.-W. (2022). SEmbedNet: Hardware-Friendly CNN for Ectopic Beat Classification on STM32-Based Edge Device. 2022 IEEE International Conference on Recent Advances in Systems Science and Engineering (RASSE), 1–6. https://doi.org/10.1109/RASSE54974.2022.9989708 [Google Scholar] [Crossref]

15. Zhang, X., Zhou, X., Lin, M., & Sun, J. (2018). ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. https://doi.org/10.1109/CVPR.2018.00716 [Google Scholar] [Crossref]

• The Role of Artificial Intelligence in Revolutionizing Library Services in Nairobi: Ethical Implications and Future Trends in User Interaction
• ESPYREAL: A Mobile Based Multi-Currency Identifier for Visually Impaired Individuals Using Convolutional Neural Network
• Comparative Analysis of AI-Driven IoT-Based Smart Agriculture Platforms with Blockchain-Enabled Marketplaces
• AI-Based Dish Recommender System for Reducing Fruit Waste through Spoilage Detection and Ripeness Assessment
• SEA-TALK: An AI-Powered Voice Translator and Southeast Asian Dialects Recognition