.
Evaluasi Keandalan Model Rekognisi Suara Burung Hama Menggunakan Platform Edge Impulse Pada Mikrokontroller Low Power
Abstract
Penelitian ini mengekplorasi kemungkinan pemanfaatan teknologi edge machine learning dalam hal rekognisi suara-suara burung hama yang bisa diaplikasikan pada mikrokontroller ultra low power. Dalam paper ini dilakukan uji kehandalan dari tiga algoritma mesin pembelajaran (ML), kemudian menyematkankannya ke mikrokontroller Seeed Xiao NRF52840 Sense. Model pembelajaran mesin yang pertama adalah Fast Convolutional Neural Netywork (CNNs) 1D dengan 2 layer, model ke-2 adalah menggunakan arsitektur berbasis transfer learning MobileNet. Dalam melakukan training dan testing digunakan mesin pembelajaran embedded platform Edge Impulse. Model pembelajaran yang dihasilkan kemudian diimplementasi sebagai Arduino Library baik sebagai representasi 32-bit floating point dan 8-bit fixed integer. Nilai dugaan yang dihasilkan oleh mikrokontroller dievaluasi dalam 4 kasus, yaitu menggunakan kompiler Edge Impulse EON dan Tensor Flow Lite (TFLite). Hasil penelitian juga melaporkan memory footprint ( RAM dan Flash), nilai akurasi, dan waktu dugaan (time inference).
Keywords
Full Text:
PDF (Bahasa Indonesia)References
Muhammad Sulton Bana, Diana Rahmawati, Koko Joni, and Miftachul Ulum, “Rancang Bangun Alat Pengusir Tikus dan Burung pada Tanaman Padi,” J-Eltrik, vol. 2, no. 1, p. 53, Nov. 2021, doi: 10.30649/j-eltrik.v2i1.53.
S. F. Nabilah, R. Agustin, and F. N. Fauziah, “Sistem Monitoring Pengusir Hama Burung Pada Tanaman Padi Menggunakan Sensor PIR dan ESPCamera Berbasis Internet Of Thing,” Politeknik Harapan Tegal, no. Tugas Akhir Teknik Komputer, pp. 1–9, 2020.
A. Muminov, Y. C. Jeon, D. Na, C. Lee, and H. S. Jeon, “Development of a solar powered bird repeller system with effective bird scarer sounds,” International Conference on Information Science and Communications Technologies, ICISCT 2017, vol. 2017-Decem, pp. 1–4, 2017, doi: 10.1109/ICISCT.2017.8188587.
R. R. Prasanna, P. Chowdary Kakarla, V. P. Simha, and N. Mohan, “IMPLEMENTATION OF TINY MACHINE LEARNING MODELS ON ARDUINO 33-BLE FOR GESTURE AND SPEECH RECOGNITION”.
W. Yu and C. Zhao, “Broad Convolutional Neural Network Based Industrial Process Fault Diagnosis with Incremental Learning Capability,” IEEE, 2020.
J.Song, “7.1 An 11.5TOPS/W 1024-MAC butterfly structure dualcore sparsity-aware neural processing unit in 8nm flagship mobile SoC,” IEEE, 2019.
B. Li, M. H. Najafi, B. Yuan, and D. J. Lilja, “Quantized neural networks with new stochastic multipliers,” IEEE, 2018.
M. H. Najafi, D. J. Lilja, and M. Riedel, “Deterministic methods for stochastic computing using low-discrepancy sequences,” IEEE, 2018.
L. Yang, W. Jiang, W. Liu, and H. Edwin, “Co-exploring neural architecture and network-on-chip design for real-time artificial intelligence,” IEEE, 2020.
W. Dai, C. Dai, S. Qu, J. Li, and S. Das, “07952190,” ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, pp. 421–425, 2017.
A. Saad, J. Ahmed, and A. Elaraby, “Classification of Bird Sound Using High-and Low-Complexity Convolutional Neural Networks,” Traitement du Signal, vol. 39, no. 1, pp. 187–193, 2022, doi: 10.18280/ts.390119.
H. , Goëau, H. , Glotin, W. P. , Vellinga, Planqué, and J. A. R., “Lifeclef bird identification task 2016: The arrival of deep learning,” in CLEF: Conference and Labs of the Evaluation Forum, 1609, https://hal.archives-ouvertes.fr/hal-01373779, 2016, pp. 440–449. [Online]. Available: https://www.edgeimpulse.com/
“Edge Impulse.” https://www.edgeimpulse.com/ (accessed Sep. 18, 2023).
J. Jongboom, “Introducing EON: Neural networks in up to 55% less RAM and 35% less ROM,” Sep. 2020. [Online]. Available: https://www.edgeimpulse.com/blog/introducing-eon
A. Saad, J. Ahmed, and A. Elaraby, “Classification of Bird Sound Using High-and Low-Complexity Convolutional Neural Networks,” Traitement du Signal, vol. 39, no. 1, pp. 187–193, 2022, doi: 10.18280/ts.390119.
Institute of Electrical and Electronics Engineers, “Bird Sound Recognition Using a Convolutional Neural Network,” in SISY 2018 • IEEE 16th International Symposium on Intelligent Systems and Informatics, Sep. 2018.
DOI: https://doi.org/10.21107/triac.v10i2.22448
Refbacks
- There are currently no refbacks.
Copyright (c) 2023 Jurnal Teknik Elektro dan Komputer TRIAC
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
