{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T14:19:33Z","timestamp":1778681973496,"version":"3.51.4"},"reference-count":46,"publisher":"MDPI AG","issue":"5","license":[{"start":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T00:00:00Z","timestamp":1778630400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Computation"],"abstract":"<jats:p>This paper investigates the design and evaluation of compact convolutional neural networks (CNNs) for keyword spotting (KWS) and acoustic event detection under the stringent constraints of the TinyML paradigm. The research expands upon traditional binary classification approaches by addressing a multi-class acoustic scenario encompassing eight distinct categories: stop, no, go, yes, unknown, silence, noise_ambient, and noise_sudden. The primary objective is to evaluate the feasibility of deploying reliable acoustic detection systems on ultra-low-power microcontrollers for edge computing applications. To this end, five lightweight architectures were developed and benchmarked: AlexNet-Tiny, LeNet-Tiny, MobileNet-Tiny, VGG-Tiny, and CustomCNN-Tiny. The models were trained using Mel-spectrogram features and optimized through INT8 post-training quantization to facilitate embedded deployment. Hardware simulation was conducted targeting the XIAO nRF52840 Sense microcontroller (64 MHz, 256 KB RAM). Experimental results demonstrate that the Gold VGG-Tiny architecture achieves the highest classification accuracy (89.81%), while Silver MobileNet-Tiny provides the superior operational efficiency with the lowest inference latency (0.88 ms) and minimal energy consumption (14.4 \u00b5J). Furthermore, the Bronze CustomCNN-Tiny model achieves the most reduced memory footprint (42.9 KB), highlighting its suitability for memory-constrained environments. Statistical validation using Cohen\u2019s Kappa, Matthews Correlation Coefficient (MCC), and Area Under the Curve (AUC) confirms the robustness and reliability of the proposed models. The potential application of this system is motivated by acoustic monitoring for the early detection of high-risk situations, such as gender-based violence. Future work will focus on on-device physical validation and real-world deployment in wearable safety electronics.<\/jats:p>","DOI":"10.3390\/computation14050112","type":"journal-article","created":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T12:46:55Z","timestamp":1778676415000},"page":"112","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Design and Evaluation of a Compact VGG-Inspired CNN for Keyword Spotting in Resource-Constrained TinyML Systems"],"prefix":"10.3390","volume":"14","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3231-0153","authenticated-orcid":false,"given":"Wilson Gustavo","family":"Chango","sequence":"first","affiliation":[{"name":"Facultad de Ingenier\u00eda en Sistemas, Electr\u00f3nica e Industrial, Universidad T\u00e9cnica de Ambato, UTA, Av. los Ch\u00e1squis, Ambato 180206, Ecuador"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3550-7173","authenticated-orcid":false,"given":"Mayra","family":"Barrera","sequence":"additional","affiliation":[{"name":"Facultad de Ciencias Humanas y de la Educaci\u00f3n, Universidad T\u00e9cnica de Ambato, UTA, Av. los Ch\u00e1squis, Ambato 180206, Ecuador"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2509-5929","authenticated-orcid":false,"given":"Daniel","family":"Maldonado-Ruiz","sequence":"additional","affiliation":[{"name":"Facultad de Ingenier\u00eda en Sistemas, Electr\u00f3nica e Industrial, Universidad T\u00e9cnica de Ambato, UTA, Av. los Ch\u00e1squis, Ambato 180206, Ecuador"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0008-6991-7139","authenticated-orcid":false,"given":"Julio","family":"Balarezo","sequence":"additional","affiliation":[{"name":"Facultad de Ingenier\u00eda en Sistemas, Electr\u00f3nica e Industrial, Universidad T\u00e9cnica de Ambato, UTA, Av. los Ch\u00e1squis, Ambato 180206, Ecuador"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7138-3913","authenticated-orcid":false,"given":"Marcelo V.","family":"Garcia","sequence":"additional","affiliation":[{"name":"Facultad de Ingenier\u00eda en Sistemas, Electr\u00f3nica e Industrial, Universidad T\u00e9cnica de Ambato, UTA, Av. los Ch\u00e1squis, Ambato 180206, Ecuador"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1069-4574","authenticated-orcid":false,"given":"Geovanny","family":"Silva","sequence":"additional","affiliation":[{"name":"Faculty of Informatics and Electronics, Escuela Superior Polit\u00e9cnica de Chimborazo (ESPOCH), Riobamba 060155, Ecuador"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2026,5,13]]},"reference":[{"key":"ref_1","first-page":"18","article-title":"Artificial Intelligence-Augmented Edge Computing: Architectures, Challenges, and Future Directions","volume":"14","author":"Rigneault","year":"2025","journal-title":"Int. J. Eng. Invent."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"591","DOI":"10.1109\/COMST.2022.3218527","article-title":"Distributed Artificial Intelligence Empowered by End-Edge-Cloud Computing: A Survey","volume":"25","author":"Duan","year":"2023","journal-title":"IEEE Commun. Surv. Tutor."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"204513","DOI":"10.1109\/ACCESS.2025.3633575","article-title":"A Systematic Review of State-of-the-Art TinyML Applications in Healthcare, Education, and Transportation","volume":"13","author":"Yahyati","year":"2025","journal-title":"IEEE Access"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"10654","DOI":"10.1109\/ACCESS.2026.3654275","article-title":"Training-Free Proxy-Guided Bayesian NAS for UAV-Constrained TinyML","volume":"14","author":"Yadlapalli","year":"2026","journal-title":"IEEE Access"},{"key":"ref_5","doi-asserted-by":"crossref","unstructured":"Guan, Y., Qiang, X., Pan, Z., Johnson, D., Fang, Y., Zhou, K., Wang, Y., Li, W., Ding, Y., and Aziz, A. (2025). Mercury: Unlocking Multi-GPU Operator Optimization for LLMs via Remote Memory Scheduling. Proceedings of the 31st ACM SIGOPS Symposium on Operating Systems Principles (SOSP \u201925), ACM.","DOI":"10.1145\/3731569.3764798"},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"Shen, J., Cheng, X., Yang, X., Zhang, L., Cheng, W., and Lin, Y. (2025). Efficient CNN Accelerator Based on Low-End FPGA with Optimized Depthwise Separable Convolutions and Squeeze-and-Excite Modules. AI, 6.","DOI":"10.3390\/ai6100244"},{"key":"ref_7","unstructured":"Hasan, M.A., and Dey, K. (2024). Depthwise Separable Convolutions with Deep Residual Convolutions. arxiv."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Rattanasak, A., Jumphoo, T., Kokkhunthod, K., Pathonsuwan, W., Nualsri, R., Thanonklang, S., Tongdee, P., Nimkuntod, P., Uthansakul, M., and Uthansakul, P. (2025). Lightweight Deep Learning Approaches on Edge Devices for Fetal Movement Monitoring. Biosensors, 15.","DOI":"10.3390\/bios15100662"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Chen, W., Wang, J., Zhang, Y., Shi, Y., Jiang, T., Zhou, S., Wu, C., Zhang, A., Zhou, C., and Wang, M. (2025). On-Device Large Language Models: A Survey of Model Compression and System Optimization. Res. Sq., Preprint.","DOI":"10.21203\/rs.3.rs-7975734\/v1"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"10498","DOI":"10.1109\/JIOT.2026.3653698","article-title":"The Internet of Humanoids: A Survey of Technologies, Applications, and Challenges","volume":"13","author":"Yu","year":"2026","journal-title":"IEEE Internet Things J."},{"key":"ref_11","first-page":"1933","article-title":"Overo: Sharing Private Audio Recordings","volume":"14","author":"Lim","year":"2022","journal-title":"Proc. ACM Conf. Comput. Commun. Secur."},{"key":"ref_12","doi-asserted-by":"crossref","unstructured":"Malche, T., Budhani, S., Soni, P.K., and Upadhyay, G.M. (2025). Voice-activated home automation system for IoT edge devices using TinyML. Discov. Internet Things, 5.","DOI":"10.1007\/s43926-025-00165-x"},{"key":"ref_13","doi-asserted-by":"crossref","unstructured":"Heydari, S., and Mahmoud, Q.H. (2025). Tiny Machine Learning and On-Device Inference: A Survey of Applications, Challenges, and Future Directions. Sensors, 25.","DOI":"10.3390\/s25103191"},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Trigkas, A., Piromalis, D., and Papageorgas, P. (2025). Edge Intelligence in Urban Landscapes: Reviewing TinyML Applications for Connected and Sustainable Smart Cities. Electronics, 14.","DOI":"10.3390\/electronics14142890"},{"key":"ref_15","first-page":"818","article-title":"Tiny Machine Learning (TinyML) Advancements for Intelligent Battery-Powered Iot Sensors","volume":"3","author":"Fatima","year":"2025","journal-title":"Spectr. Eng. Sci."},{"key":"ref_16","unstructured":"Roh, H., Yeo, J., Ko, Y., Wei, G.Y., Brooks, D., and Choi, W.S. (2025). Flash: A Hybrid Private Inference Protocol for Deep CNNs with High Accuracy and Low Latency on CPU. arXiv."},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Somvanshi, S., Islam, M.M., Chhetri, G., Chakraborty, R., Mimi, M.S., Shuvo, S.A., Islam, K.S., Javed, S., Rafat, S.A., and Dutta, A. (2026). From Tiny Machine Learning to Tiny Deep Learning: A Survey. ACM Comput. Surv., 58.","DOI":"10.1145\/3776588"},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"10564","DOI":"10.1109\/JIOT.2026.3656932","article-title":"A Comprehensive Survey of TinyML-Based Biometric Recognition for IoT Edge Devices","volume":"13","author":"Essahraui","year":"2026","journal-title":"IEEE Internet Things J."},{"key":"ref_19","unstructured":"Garai, S., and Samui, S. (2025). Advances in Small-Footprint Keyword Spotting: A Comprehensive Review of Efficient Models and Algorithms. arXiv."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"De Marco, R., Di Nardo, F., Rongoni, A., Screpanti, L., and Scaradozzi, D. (2025). Real-Time Dolphin Whistle Detection on Raspberry Pi Zero 2 W with a TFLite Convolutional Neural Network. Robotics, 14.","DOI":"10.3390\/robotics14050067"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"34613","DOI":"10.1109\/ACCESS.2025.3543838","article-title":"Toward Efficient Cancer Detection on Mobile Devices","volume":"13","author":"Lee","year":"2025","journal-title":"IEEE Access"},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"114145","DOI":"10.1109\/ACCESS.2025.3581271","article-title":"Customized Spectro-Temporal CNN Feature Extraction and ELM-Based Classifier for Accurate Respiratory Obstruction Detection","volume":"13","author":"Muthulakshmi","year":"2025","journal-title":"IEEE Access"},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Munirathinam, R., and Vitek, S. (2026). Hybrid and Depth Separable Convolutional Neural Network Approaches for an Efficient and Accurate Siren Recognition with Resource Limited Devices. J. Sens., 2026.","DOI":"10.1155\/js\/4353782"},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"El Zeinaty, C., Hamidouche, W., Herrou, G., and Menard, D. (2025). Designing Object Detection Models for TinyML: Foundations, Comparative Analysis, Challenges, and Emerging Solutions. ACM Comput. Surv., 58.","DOI":"10.1145\/3744339"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Bae, S., Kim, H., Lee, S., and Jung, Y. (2023). FPGA Implementation of Keyword Spotting System Using Depthwise Separable Binarized and Ternarized Neural Networks. Sensors, 23.","DOI":"10.3390\/s23125701"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Xue, Y., Ng, W.S., Goh, W.L., Gao, Y., and Lee, S. (2025, January 16\u201318). A Real-time Keyword Spotting Accelerator Using Depthwise Separable Convolutional Neural Network. Proceedings of the 2025 IEEE Biomedical Circuits and Systems Conference (BioCAS), Abu Dhabi, UAE.","DOI":"10.1109\/BioCAS67066.2025.00020"},{"key":"ref_27","first-page":"62414","article-title":"Pruning vs. Quantization: Which is Better?","volume":"36","author":"Kuzmin","year":"2023","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"139113","DOI":"10.1109\/ACCESS.2024.3465631","article-title":"Advances in Pruning and Quantization for Natural Language Processing","volume":"12","author":"Bibi","year":"2024","journal-title":"IEEE Access"},{"key":"ref_29","unstructured":"Lettiero, M., and De Luca, P. (July, January 30). Parallel Optimization of Quantized CNNs for Efficient GPU Inference. Proceedings of the International Joint Conference on Neural Networks (IJCNN), Rome, Italy."},{"key":"ref_30","unstructured":"Zhang, Y., Sun, B., Jiang, W., Ha, Y., Hu, M., and Zhao, W. (November, January 30). WSQ-AdderNet: Efficient Weight Standardization Based Quantized AdderNet FPGA Accelerator Design with High-Density INT8 DSP-LUT Co-Packing Optimization. Proceedings of the 41st IEEE\/ACM International Conference on Computer-Aided Design (ICCAD), San Diego, CA, USA."},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Khadse, S. (2025). Small Language Models and Efficient AI: The Future of Sustainable, Accessible Intelligence: A Comprehensive Analysis of Model Compression, Edge Deployment, and Resource-Efficient AI Systems. SSRN Electron. J., preprint.","DOI":"10.2139\/ssrn.5664971"},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"21362","DOI":"10.1109\/JSEN.2022.3210773","article-title":"Machine Learning for Microcontroller-Class Hardware: A Review","volume":"22","author":"Saha","year":"2022","journal-title":"IEEE Sens. J."},{"key":"ref_33","doi-asserted-by":"crossref","unstructured":"Immonen, R., and H\u00e4m\u00e4l\u00e4inen, T. (2022). Tiny Machine Learning for Resource-Constrained Microcontrollers. J. Sens., 2022.","DOI":"10.1155\/2022\/7437023"},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Shabir, M.Y., Torta, G., and Damiani, F. (2025). TinyML model compression: A comparative study of pruning and quantization on selected standard and custom neural networks. Telecommun. Syst., 88.","DOI":"10.1007\/s11235-025-01363-2"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"12167","DOI":"10.1109\/ACCESS.2026.3656854","article-title":"A Novel FastKAN with Few-Shot Learning for Real-Time Driver Distraction Detection on TinyML Microcontrollers","volume":"14","author":"Yahyati","year":"2026","journal-title":"IEEE Access"},{"key":"ref_36","unstructured":"Mitsis, S., Hadjikyriakos, E., Ibrahim, H., Neofytou, S., Raman, S., Myles, J., and Kanjo, E. (2025). Transformer Redesign for Late Fusion of Audio-Text Features on Ultra-Low-Power Edge Hardware. arXiv."},{"key":"ref_37","doi-asserted-by":"crossref","unstructured":"Vicente, E.C., Silva, L.A., da Rocha Fernandes, A.M., and Parreira, W.D. (2025). A Structured Review of IoT-Based Embedded Systems and Machine Learning for Water Quality Monitoring. Appl. Sci., 15.","DOI":"10.3390\/app152111719"},{"key":"ref_38","doi-asserted-by":"crossref","unstructured":"Kumar V, S., Nadh B, R., S, S., Suresh Kumar, A., Raj S R, A., and G, S. (2025). Leveraging Artificial Neural Networks for Real-Time Speech Recognition in Voice-Activated Systems. ITM Web Conf., 76.","DOI":"10.1051\/itmconf\/20257601003"},{"key":"ref_39","doi-asserted-by":"crossref","unstructured":"He, Y., Tao, Y., and An, L. (2023, January 29\u201331). Real-Time Voice Command Recognition System Based on DTW and MFCC. Proceedings of the 6th International Conference on Computer Network, Electronic and Automation (ICCNEA), Xi\u2019an, China.","DOI":"10.1109\/ICCNEA60107.2023.00084"},{"key":"ref_40","doi-asserted-by":"crossref","unstructured":"Bhroge, S., Mane, V., Wyawahare, M., Kamble, M., Rane, M., Sapate, R., and Raut, S. (2025). Hybrid CNN-BiLSTM for Deepfake Voice Detection: A Comparative Study. Int. J. Saf. Secur. Eng., 15.","DOI":"10.18280\/ijsse.150907"},{"key":"ref_41","doi-asserted-by":"crossref","unstructured":"Jiang, J., Wen, B., Ma, J., Fang, L., Jiang, K., Fei, X., Wu, S., Miao, J., Wang, G., and Sun, L. (2024). Lightweight-VGG: A Fast Deep Learning Architecture Based on Dimensionality Reduction and Nonlinear Enhancement for Hyperspectral Image Classification. Remote Sens., 16.","DOI":"10.3390\/rs16020259"},{"key":"ref_42","first-page":"1","article-title":"Resource-Efficient Neural Networks for Embedded Systems","volume":"25","author":"Roth","year":"2024","journal-title":"J. Mach. Learn. Res."},{"key":"ref_43","doi-asserted-by":"crossref","unstructured":"L\u00ea, M.T., Wolinski, P., and Arbel, J. (2026). Efficient Neural Networks for Tiny Machine Learning: A Comprehensive Review. ACM Trans. Intell. Syst. Technol., 17.","DOI":"10.1145\/3798276"},{"key":"ref_44","unstructured":"Fanariotis, A., Orphanoudakis, T., and Fotopoulos, V. (2025). Evaluating the Energy Efficiency of NPU-Accelerated Machine Learning Inference on Embedded Microcontrollers. arXiv."},{"key":"ref_45","doi-asserted-by":"crossref","first-page":"55756","DOI":"10.1109\/JIOT.2025.3625087","article-title":"An Explainable Tiny-Fast Kolmogorov-Arnold Network for Gesture-Based Air Handwriting Recognition of Tifinagh Letters in Resource-Constrained IoT Device","volume":"12","author":"Lamaakal","year":"2025","journal-title":"IEEE Internet Things J."},{"key":"ref_46","doi-asserted-by":"crossref","first-page":"567","DOI":"10.1111\/2041-210x.70206","article-title":"Leveraging machine learning and accelerometry to classify animal behaviours with uncertainty","volume":"17","author":"Agarwal","year":"2025","journal-title":"Methods Ecol. Evol."}],"container-title":["Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2079-3197\/14\/5\/112\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,13]],"date-time":"2026-05-13T13:43:26Z","timestamp":1778679806000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2079-3197\/14\/5\/112"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,13]]},"references-count":46,"journal-issue":{"issue":"5","published-online":{"date-parts":[[2026,5]]}},"alternative-id":["computation14050112"],"URL":"https:\/\/doi.org\/10.3390\/computation14050112","relation":{},"ISSN":["2079-3197"],"issn-type":[{"value":"2079-3197","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,13]]}}}