Implementasi sistem kecerdasan buatan berbasis deep learning yang dilakukan pada perangkat keras standar mengalami dua hambatan utama: beban komputasi tinggi dan kompleksitas pengelolaan dependensi perangkat lunak lintas platform. Penelitian ini bertujuan membangun sistem deteksi emosi wajah real-time yang ringan dan mudah dengan menggunakan TensorFlow Lite (TFLite) dan Docker Container. Model Convolutional Neural Network (CNN) dilatih pada dataset FER-2013 (28.709 citra latih, 7.178 citra uji, 7 kelas emosi) kemudian dikompresi ke format TFLite. Evaluasi menghasilkan akurasi keseluruhan 51,94% dengan kelas Happy mencapai F1-Score tertinggi (72,73%). Latensi inferensi rata-rata 0,309 ms per citra dengan throughput 3.241 FPS membuktikan kemampuan real-time yang melampaui standar 24 FPS. Kompresi TFLite mereduksi ukuran model 97% (50–100 MB menjadi 1,4 MB). Sistem berhasil dieksekusi pada laptop standar dalam lingkungan Docker tanpa akselerator perangkat keras khusus.
References
[1] E. Sariyanidi, H. Gunes, and A. Cavallaro, "Automatic analysis of facial affect: A survey of registration, representation, and recognition," IEEE Trans. Pattern Anal. Mach. Intell., vol. 37, no. 6, pp. 1113–1133, Jun. 2015, doi: 10.1109/TPAMI.2014.2366127.
[2] D. Sculley et al., "Hidden technical debt in machine learning systems," in Proc. Adv. Neural Inf. Process. Syst. (NeurIPS), 2015, vol. 28, pp. 2503–2511.
[3] I. J. Goodfellow et al., "Challenges in representation learning: A report on three machine learning contests," Neural Netw., vol. 64, pp. 59–63, Apr. 2015, doi: 10.1016/j.neunet.2014.09.005.
[4] C. Boettiger, "An introduction to Docker for reproducible research," ACM SIGOPS Oper. Syst. Rev., vol. 49, no. 1, pp. 71–79, Jan. 2015, doi: 10.1145/2723872.2723882.
[5] R. David et al., "TensorFlow Lite Micro: Embedded machine learning for TinyML systems," Proc. Mach. Learn. Syst. (MLSys), vol. 3, pp. 800–811, 2021.
[6] D. Merkel, "Docker: Lightweight Linux containers for consistent development and deployment," Linux J., vol. 2014, no. 239, Art. 2, Mar. 2014.
[7] Y. LeCun, Y. Bengio, and G. Hinton, "Deep learning," Nature, vol. 521, no. 7553, pp. 436–444, May 2015, doi: 10.1038/nature14539.
[8] P. Viola and M. Jones, "Rapid object detection using a boosted cascade of simple features," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2001, vol. 1, pp. I-511–I-518, doi: 10.1109/CVPR.2001.990517.
[9] P. Ekman and W. V. Friesen, "Constants across cultures in the face and emotion," J. Pers. Soc. Psychol., vol. 17, no. 2, pp. 124–129, 1971, doi: 10.1037/h0030377.
[10] C. Pramerdorfer and M. Kampel, "Facial expression recognition using convolutional neural networks: State of the art," arXiv:1612.02903, Dec. 2016. [Online]. Available: https://arxiv.org/abs/1612.02903
[11] A. Howard et al., "MobileNets: Efficient convolutional neural networks for mobile vision applications," arXiv:1704.04861, Apr. 2017. [Online]. Available: https://arxiv.org/abs/1704.04861
[12] A. Wibowo, A. Santoso, and H. Nugroho, "Implementasi TensorFlow Lite untuk klasifikasi gestur tangan pada Raspberry Pi berbasis kamera," J. Tek. Inform. dan Sist. Inf., vol. 8, no. 2, pp. 145–156, 2022.
[13] B. Prasetyo and L. Santoso, "Optimasi deteksi objek real-time menggunakan TFLite pada sistem pengawasan bertenaga baterai," J. Nas. Tek. Elektro Teknol. Inf. (JNTETI), vol. 12, no. 1, pp. 45–53, 2023.
[14] S. Li and W. Deng, "Deep facial expression recognition: A survey," IEEE Trans. Affect. Comput., vol. 13, no. 3, pp. 1195–1215, 2022, doi: 10.1109/TAFFC.2020.2981446.
[15] Z. Zhao, Q. Liu, and F. Zhou, "Robust lightweight facial expression recognition network with label distribution training," in Proc. AAAI Conf. Artif. Intell., 2021, vol. 35, no. 4, pp. 3510–3519, doi: 10.1609/aaai.v35i4.16467.