Performance Evaluation and Comparison of YOLO11 Model Variants for Indonesian Sign Language (BISINDO)

Dimas Eka Putra Sahtio; Made Adi Paramartha Putra; Ida Bagus Kresna Sudiatmika

doi:10.23960/jitet.v14i2.9348

Dimas Eka Putra Sahtio
Primakara University
Made Adi Paramartha Putra
Primakara University
Ida Bagus Kresna Sudiatmika
Primakara University

DOI: https://doi.org/10.23960/jitet.v14i2.9348

Keywords BISINDO, Deep Learning, Hyperparameter Tuning, Object Detection, YOLO11

Abstract Views (Last 12 Months)

83 Abstract Views

48 Downloads

Abstract

This study aims to evaluate and compare the performance of various versions of the You Only Look Once (YOLO) versionx 11 model in detecting words in Indonesian Sign Language (BISINDO). Communication is a fundamental need, but individuals with hearing impairments often face challenges due to the public's lack of understanding of sign language. The use of Artificial Intelligence (AI) with a Deep Learning approach, specifically the YOLO algorithm, provides a solution for automatically translating sign language. Although YOLOv8 has been used previously, the update to YOLO11 promises improvements in efficiency and accuracy. In this study, hyperparameter tuning was conducted to obtain the best configuration (batch size 16, lr0=0.001, lrf=0.1), which was then used to train and compare YOLO11 variants (YOLO11n, YOLO11s, YOLO11m) on a dataset of 36 BISINDO classes. The results showed that the YOLO11n model provided the most optimal performance in terms of balancing accuracy and time efficiency, with an mAP50-95 value of 88.7%, a training duration of 3.771 hours, and an inference speed of 3.6 ms. The YOLO11s variant had a slightly higher detection accuracy with an mAP50-95 of 89.2%, with 5.351 inference time.

Downloads

Download data is not yet available.

References

N. B. A. Karna, M. A. P. Putra, S. M. Rachmawati, M. Abisado and G. A. Sampedro, "Toward Accurate Fused Deposition Modeling 3D Printer Fault Detection Using Improved YOLOv8 With Hyperparameter Optimization," in IEEE Access, vol. 11, pp. 74251-74262, 2023, doi: 10.1109/ACCESS.2023.3293056.

N. S. W. Nugroho and M. P. K. Putra, "Leveraging deep learning approach for accurate alphabet recognition through hand gestures in sign language," Jurnal Teknik Informatika (JUTIF), vol. 6, 2025.

D. S. Ariansyah, "Pendeteksi Kata dalam Bahasa Isyarat Menggunakan Algoritma YOLO Versi 8," Jurnal Informatika dan Teknik Elektro Terapan, vol. 12, no. 3, 2024, doi: 10.23960/jitet.v12i3.4904.

A. Pangestu, M. Muttaqin, and M. Sunandar, "Sistem Deteksi Bahasa Isyarat Indonesia (Bisindo) Menggunakan Algoritma You Only Look Once (YOLO)v8," JATI (Jurnal Mahasiswa Teknik Informatika), vol. 8, no. 5, pp. 9891–9897, 2024.

N. Navin et al., "Bilingual Sign Language Recognition: A YOLOv11-Based Model for Bangla and English Alphabets," Journal of Imaging, vol. 11, no. 5, p. 134, 2025, doi: 10.3390/jimaging11050134.

R. Raj, R. Sreemathy, M. Turuk, J. Jagdale, and M. Anish, "Performance Comparison of Different Versions of YOLO for Indian Sign Language Captioning in Real Time of Multiple Signers," Procedia Computer Science, vol. 259, pp. 991–1000, 2025, doi: 10.1016/j.procs.2025.04.053.

A. Y. Pardede, "BISINDO 40 Kata mp4," 2024. [Online]. Available: https://www.kaggle.com/datasets/anggiyohanespardede/bisindo-40-kata-mp4

G. Wiriasto, A. Rizaldy, P. Wiguna, and I. Kinasih, "Drip Infusion Monitoring and Data Logging System Based on YOLOv5," JITK (Jurnal Ilmu Pengetahuan Dan Teknologi Komputer), vol. 11, pp. 171–179, 2025.

T. Diwan, G. Anirudh, and J. V. Tembhurne, "Object detection using YOLO: challenges, architectural successors, datasets and applications," Multimedia Tools and Applications, vol. 82, no. 6, pp. 9243–9275, 2023, doi: 10.1007/s11042-022-13644-y.

P. Hidayatullah, N. Syakrani, M. R. Sholahuddin, T. Gelar, and R. Tubagus, "YOLOv8 to YOLO11: A Comprehensive Architecture In-depth Comparative Review," 2025. [Online]. Available: https://arxiv.org/pdf/2501.13400.

R. Supriyadi et al., "Pengembangan Aplikasi Estimasi Kalori Makanan Berbasis Citra dengan Pendekatan Deteksi Objek Menggunakan YOLO," Jurnal Informatika dan Teknik Elektro Terapan, vol. 14, no. 1, 2026, doi: 10.23960/jitet.v14i1.8545.

K. Wang, J. Liu, and X. Cai, "C2PSA-enhanced YOLOv11 architecture: A novel approach for small target detection in cotton disease diagnosis," arXiv preprint arXiv:2508.12219, 2025.

D.-L. Nguyen, X.-T. Vo, A. Priadana, J. Choi, and K.-H. Jo, "Improved YOLOv11 Based on Attention and Extra Head Modules for Dental Disease Detection," in 2025 International Workshop on Intelligent Systems (IWIS), Ulsan, South Korea, 2025, pp. 1–6, doi: 10.1109/IWIS66215.2025.11142416.

L. Quaranta, F. Calefato, and F. Lanubile, "KGTorrent: A Dataset of Python Jupyter Notebooks from Kaggle," in 2021 IEEE/ACM 18th International Conference on Mining Software Repositories (MSR), Madrid, Spain, 2021, pp. 550–554, doi: 10.1109/MSR52588.2021.00072.

M. A. P. Putra, I. Utama, N. W. Utami, and I. G. E. J. Putra, "Enhancing Federated Learning Performance through Adaptive Client Optimization with Hyperparameter Tuning," Journal of Applied Data Sciences, vol. 5, no. 2, pp. 747–755, 2024, doi: 10.47738/jads.v5i2.251.