Developing Cost-Effective AI Algorithms for Resource-Constrained Devices
DOI:
https://doi.org/10.59992/IJCI.2025.v4n1p3Keywords:
Resource-Constrained Devices, Cost-Effective AI, Model Pruning, Quantization, Energy Efficiency, Edge ComputingAbstract
Resource-restrained gadgets pose sizable challenges for deploying synthetic intelligence (AI) packages, which include restricted computational power, reminiscence, and electricity resources. This research pursuits to broaden value-effective AI algorithms that cope with these limitations whilst retaining high overall performance and accuracy. The examine leverages superior optimization strategies, such as version pruning, quantization, and dynamic strength control, to layout light-weight models appropriate for low-strength environments.
Experiments conducted on gadgets like the Raspberry Pi 4 and NVIDIA Jetson Nano screen giant improvements in inference time, electricity efficiency, and accuracy compared to conventional processes. The proposed algorithms acquire up to 50% reduction in energy consumption and 20% improvement in accuracy at the same time as lowering typical computational charges. These findings reveal the feasibility of deploying green AI solutions on constrained hardware without compromising on functionality or nice.
The practical implications of this paintings make bigger to various applications, along with real-time healthcare monitoring, clever agriculture, and commercial IoT systems. The have a look at concludes by means of highlighting areas for destiny studies, which includes improving algorithmic adaptability and expanding trying out to embody diverse eventualities. This work gives a sturdy basis for advancing the deployment of AI in resource-restricted settings, bridging the gap between technological innovation and practical implementation.
References
1. Canepa, A. (2023). Application-aware optimization of Artificial Intelligence for deployment on resource constrained devices.
2. Nasir, M., Muhammad, K., Ullah, A., Ahmad, J., Baik, S. W., & Sajjad, M. (2022). Enabling automation and edge intelligence over resource constraint IoT devices for smart home. Neurocomputing, 491, 494-506.
3. Neseem, M. (2024). AI at the Edge: Efficient Deep Learning for Resource-Constrained Environments (Doctoral dissertation, Brown University PROVIDENCE, RHODE ISLAND).
4. 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.
5. S. Wendelken and C. MacGillivray. Worldwide and U.S. IoT Cellular Connections Forecast, 2021–2025. Accessed: Feb. 17, 2022. [Online]. Available: https://www.idc.com/getdoc.jsp?containerId=US47296121
6. M. Capra, B. Bussolino, A. Marchisio, M. Shafique, G. Masera, and M. Martina, “An updated survey of efficient hardware architectures for accelerating deep convolutional neural networks,” Future Internet, vol. 12, no. 7, p. 113, Jul. 2020.
7. S. Dave, R. Baghdadi, T. Nowatzki, S. Avancha, A. Shrivastava, and B. Li, “Hardware acceleration of sparse and irregular tensor computations of ML models: A survey and insights,” Proc. IEEE, vol. 109, no. 10, pp. 1706–1752, Oct. 2021.
8. P. Refaeilzadeh, L. Tang, H. Liu, L. Angeles, and C. D. Scientist, “Cross-validation,” Encyclopedia Database Syst., vol. 5, pp. 532–538, Jan. 2020.
9. J. E. Van Engelen and H. H. Hoos, “A survey on semi-supervised learning,” Mach. Learn., vol. 109, no. 2, pp. 373–440, Feb. 2020.
10. X. Zhang, X. Zhou, M. Lin, and J. Sun, “ShuffleNet: An extremely efficient convolutional neural network for mobile devices,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit., Jun. 2018, pp. 6848–6856.
11. Z. Jelˇcicová and M. Verhelst, “Delta keyword transformer: Bringing transformers to the edge through dynamically pruned multi-head self-attention,” 2022, arXiv: 2204.03479.
12. Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott E. Reed, Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, and Andrew Rabi- novich. Going deeper with convolutions. CoRR, abs/1409.4842, 2014. URL http://arxiv.org/abs/1409.4842.
13. Edoardo Ragusa, Christian Gianoglio, Rodolfo Zunino, and Paolo Gastaldo. Image polarity detection on resource-constrained devices. IEEE Intelligent Systems, 2020.
14. Hu Lin, Jingkai Zhou, Yanfen Gan, Chi-Man Vong, and Qiong Liu. Novel up-scale feature aggregation for object detection in aerial images. Neurocomputing, 411: 364–374, 2020.
15. Edoardo Ragusa, Christian Gianoglio, Filippo Dalmonte, and Paolo Gastaldo. Video grasping classification enhanced with automatic annotations. In International Conference on Applications in Electronics Pervading Industry, Environment and Society, pages 23–29. Springer, 2020.
16. Mate Kriˇsto, Marina Ivasic-Kos, and Miran Pobar. Thermal object detection in difficult weather conditions using yolo. IEEE Access, 8:125459–125476, 2020.
17. Linjie Xing, Xiaoyan Fan, Yaxin Dong, Zenghui Xiong, Lin Xing, Yang Yang, Haicheng Bai, and Chengjiang Zhou. Multi-uav cooperative system for search and rescue based on yolov5. International Journal of Disaster Risk Reduction, 76: 102972, 2022.
18. Lucas Prado Osco, Jos´e Marcato Junior, Ana Paula Marques Ramos, L´ucio Andr´e de Castro Jorge, Sarah Narges Fatholahi, Jonathan de Andrade Silva, Ed- son Takashi Matsubara, Hemerson Pistori, Wesley Nunes Gon¸calves, and Jonathan Li. A review on deep learning in uav remote sensing. International Journal of Applied Earth Observation and Geoinformation, 102:102456, 2021.
19. M. M. H. Shuvo, O. Hassan, D. Parvin, M. Chen, and S. K. Islam, “An optimized hardware
20. Implementation of deep learning inference for diabetes prediction,” in Proc. IEEE Int. Instrum. Meas. Technol. Conf. (I2MTC), May 2021, pp. 1–6.
