Superioritas Optimizer AdaGrad pada Arsitektur Spasio-Temporal PolyFace-LSTM untuk Prediksi Kepribadian Berbasis Video
-
Muhammad Zidan Alif Oktavian
*
Universitas Negeri Surabaya, Surabaya,
Indonesia
- Salamun Rohman Nudin Universitas Negeri Surabaya, Surabaya, Indonesia (*) Corresponding Author
Abstract
Conventional personality assessment generally relies on psychometric questionnaires, such as the Big Five Inventory (BFI). However, this method is susceptible to self-report bias and requires a relatively long evaluation time. On the other hand, the development of automated video-based personality analysis systems faces high computational challenges in simultaneously integrating facial spatial features and the temporal dynamics of micro-expressions. This study aims to propose a solution in the form of a hybrid spatio-temporal neural network architecture utilizing transfer learning techniques to predict personality more objectively. The proposed system integrates the Multi-Task Cascaded Convolutional Networks (MTCNN) algorithm for automatic face detection, the PolyFace model as a spatial feature extractor, and the Long Short-Term Memory (LSTM) algorithm to model temporal relationships between frames. Experiments were conducted by comparing three optimization algorithms, namely Adam, AdaGrad, and SGD, using the ChaLearn LAP 2017 dataset. The results show that the AdaGrad optimizer achieved the best generalization and prediction performance during the validation phase, with an average accuracy of 89.13%, outperforming SGD (88.34%) and Adam (88.33%). Theoretically and mathematically, the superiority of AdaGrad in this architecture is attributed to its ability to dynamically adjust the learning rate based on the accumulation of historical gradients. This mechanism makes AdaGrad considerably more stable in extracting complex spatio-temporal micro-expression features without becoming trapped in local minima. The highest performance achieved by AdaGrad was observed in the Agreeableness dimension, reaching an accuracy of 90.14%. The main contribution of this study is the development of an efficient (low-latency), precise spatio-temporal personality prediction model that avoids local minima, making it suitable for implementation in real-world automated psychological detection systems.
Downloads
References
C. J. Soto, “How Replicable Are Links Between Personality Traits and Consequential Life Outcomes? The Life Outcomes of Personality Replication Project,” Psychol. Sci., vol. 30, no. 5, hal. 711–727, 2019, doi: 10.1177/0956797619831612.
C. Suman, S. Saha, A. Gupta, S. K. Pandey, dan P. Bhattacharyya, “A multi-modal personality prediction system,” Knowledge-Based Syst., vol. 236, hal. 107–715, 2022, doi: 10.1016/j.knosys.2021.107715.
D. H. M. Pelt, D. Van der Linden, C. S. Dunkel, dan M. P. Born, “The Motivation and Opportunity for Socially Desirable Responding Does Not Alter the General Factor of Personality,” Assessment, vol. 28, no. 5, hal. 1376–1396, 2021, doi: 10.1177/1073191119880960.
T. Yan dan D. Williams, “Response Burden - Review and Conceptual Framework,” J. Off. Stat., vol. 38, no. 4, hal. 939–961, 2022, doi: 10.2478/jos-2022-0041.
S. Afzal, H. Ali Khan, M. Jalil Piran, dan J. Weon Lee, “A Comprehensive Survey on Affective Computing: Challenges, Trends, Applications, and Future Directions,” IEEE Access, vol. 12, no. July, hal. 96150–96168, 2024, doi: 10.1109/ACCESS.2024.3422480.
Y. Li, J. Wei, Y. Liu, J. Kauttonen, dan G. Zhao, “Deep Learning for Micro-Expression Recognition: A Survey,” IEEE Trans. Affect. Comput., vol. 13, no. 4, hal. 2028–2046, 2022, doi: 10.1109/TAFFC.2022.3205170.
A. Arora dan S. A. Dhondiyal, “Advancements in Face Detection: A Comparative Study of Multitask Cascaded Convolutional Networks (MTCNN) and Leading Algorithms,” 2026, hal. 531–554. doi: 10.1007/978-981-96-7140-3_36.
S. Minaee, A. Abdolrashidi, H. Su, M. Bennamoun, dan D. Zhang, “Biometrics recognition using deep learning: a survey,” Artif. Intell. Rev., vol. 56, no. 8, hal. 8647–8695, 2023, doi: 10.1007/s10462-022-10237-x.
R. M. Schmidt, F. Schneider, dan P. Hennig, “Descending through a Crowded Valley - Benchmarking Deep Learning Optimizers,” Proc. Mach. Learn. Res., vol. 139, hal. 9367–9376, 2021.
S. Aslan, U. Güdükbay, dan H. Dibeklioğlu, “Multimodal assessment of apparent personality using feature attention and error consistency constraint,” Image Vis. Comput., vol. 110, 2021, doi: 10.1016/j.imavis.2021.104163.
Y. Liu, G. Song, M. Zhang, J. Liu, Y. Zhou, dan J. Yan, “Towards flops-constrained face recognition,” Proc. - 2019 Int. Conf. Comput. Vis. Work. ICCVW 2019, hal. 2698–2702, 2019, doi: 10.1109/ICCVW.2019.00330.
H. J. Escalante dkk., “Modeling, Recognizing, and Explaining Apparent Personality from Videos,” IEEE Trans. Affect. Comput., vol. 13, no. 2, hal. 894–911, 2022, doi: 10.1109/TAFFC.2020.2973984.
A. Ouarka, T. A. Baha, Y. Es-Saady, dan M. El Hajji, “A deep multimodal fusion method for personality traits prediction,” Multimed. Tools Appl., no. 0123456789, 2024, doi: https://doi.org/10.1007/s11042-024-20356-y.
K. D. Anggara, D. P. Kartikasari, dan F. A. Bakhtiar, “Implementasi Algoritma MTCNN dalam Mekanisme Autentikasi berbasis Pengenalan Wajah,” J. Pengemb. Teknol. Inf. dan Ilmu Komput., vol. 7, no. 8, hal. 3613–3621, 2023.
A. Géron, Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow. O’Reilly Media, 2022.
J. C. S. J. Junior, A. Lapedriza, C. Palmero, X. Baro, dan S. Escalera, “Person Perception Biases Exposed: Revisiting the First Impressions Dataset,” Proc. - 2021 IEEE Winter Conf. Appl. Comput. Vis. Work. WACVW 2021, hal. 13–21, 2021, doi: 10.1109/WACVW52041.2021.00006.
A. Nogales, J. M. Sánchez Velázquez, M. Cai, A. Coney, dan Á. J. García-Tejedor, “Lightweight hybrid spatiotemporal deep learning models for video frame prediction: A comparative study across dataset complexities,” Image Vis. Comput., vol. 174, hal. 106116, Okt 2026, doi: 10.1016/j.imavis.2026.106116.
A. R. Bhafiel, A. Junaidi, M. Muharrom, dan A. Haromainy, “Analisis perbandingan akurasi algoritma haar cascade dan MTCNN untuk deteksi wajah pada kondisi cahaya Comparative Analysis of the Accuracy of the Haar Cascade and MTCNN Algorithms for Face Detection in Indoor Lighting Conditions,” Pros. Semin. Nas. Penelit. dan Pengabdi. Kpd. Masy. LPPM Univ. ‘Aisyiyah Yogyakarta, vol. 4, hal. 56–62, 2026.
S. R. Rayarao dan N. Donikena, “PyTorch: Revolutionizing Deep Learning Through Dynamic Computation Graphs and Pythonic Design,” arXiv, 2025, [Daring]. Tersedia pada: https://www.authorea.com/users/868916/articles/1324488-pytorch-revolutionizing-deep-learning-through-dynamic-computation-graphs-and-pythonic-design?commit=5815706934806f8bd5c2053b246f5f8805063869
T. Pham, A. Furno, F. Chamroukhi, dan L. Oukhellou, “Federated dynamic modeling and learning for spatiotemporal data forecasting,” Neurocomputing, vol. 672, hal. 132712, Apr 2026, doi: 10.1016/j.neucom.2026.132712.
S. Hochreiter dan J. Schmidhuber, “Long Short-Term Memory,” Neural Comput., vol. 9, no. 8, hal. 1735–1780, Nov 1997, doi: 10.1162/neco.1997.9.8.1735.
I. Salehin dan D.-K. Kang, “A Review on Dropout Regularization Approaches for Deep Neural Networks within the Scholarly Domain,” Electronics, vol. 12, no. 14, hal. 3106, Jul 2023, doi: 10.3390/electronics12143106.
Y. Shao dkk., “BDS-Adam optimizer integrating adaptive variance rectification with semi-adaptive gradient smoothing,” Sci. Rep., vol. 15, no. 1, hal. 36906, Okt 2025, doi: 10.1038/s41598-025-20788-y.
Z. Liu, “Can Adaptive Gradient Methods Converge under Heavy-Tailed Noise? A Case Study of AdaGrad,” 2026, [Daring]. Tersedia pada: http://arxiv.org/abs/2605.18694
C. R. Oktarina, Andini Setyo Anggraeni, Muhammad Arib Alwansyah, dan Reza Pahlepi, “Outlier Handling in Applied Regression: Performance Comparison Between Least Trimmed Squares and Maximum Likelihood-Type Estimators,” J-KOMA J. Ilmu Komput. dan Apl., vol. 8, no. 02, hal. 1–8, Des 2025, doi: 10.21009/JKOMA.082.01.
Y. Li, L. Gao, S. Hu, G. Gui, dan C.-Y. Chen, “Nonlocal low-rank plus deep denoising prior for robust image compressed sensing reconstruction,” Expert Syst. Appl., vol. 228, hal. 120456, Okt 2023, doi: 10.1016/j.eswa.2023.120456.
NVIDIA Corporation, “GeForce RTX 5070 Family,” NVIDIA. Diakses: 21 Agustus 2026. [Daring]. Tersedia pada: https://www.nvidia.com/id-id/geforce/graphics-cards/50-series/rtx-5070-family/](https://www.nvidia.com/id-id/geforce/graphics-cards/50-series/rtx-5070-family/
PyTorch Foundation, “Start Locally,” PyTorch. Diakses: 21 Agustus 2026. [Daring]. Tersedia pada: https://pytorch.org/get-started/locally/](https://pytorch.org/get-started/locally/
Bila bermanfaat silahkan share artikel ini
Berikan Komentar Anda terhadap artikel Superioritas Optimizer AdaGrad pada Arsitektur Spasio-Temporal PolyFace-LSTM untuk Prediksi Kepribadian Berbasis Video
Pages: 752-762
Copyright (c) 2026 Muhammad Zidan Alif Oktavian, Salamun Rohman Nudin
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under Creative Commons Attribution 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (Refer to The Effect of Open Access).
