cover
Contact Name
Arie Vatresia
Contact Email
arie.vatresia@unib.ac.id
Phone
+6282179370950
Journal Mail Official
arie.vatresia@unib.ac.id
Editorial Address
Jalan W.R. Supratman gang Cipta Baru no. 12 RT/RW 19/01 Talang Kering
Location
Kota bengkulu,
Bengkulu
INDONESIA
Jurnal Pseudocode
Published by Universitas Bengkulu
ISSN : 23555920     EISSN : 26551845     DOI : https://doi.org/10.33369
Pseudocodeis a scientific journal in the information science family that contains the results of informatics research, scientific literature on informatics, and reviews of the development of theories, methods, and application of informatics engineering science. Pseudocode is published by the Informatics Study Program, Faculty of Engineering, University of Bengkulu. Editors invite researchers, practitioners, and students to submit article manuscripts in the field of informatics engineering. Pseudocode is published 2 (two) times a year in February and September with p-ISSN 2355-5920 e-ISSN 2655-1845. Jurnal Pseudocode is Accredited by the Ministry for Research, Technology and Higher Education (RISTEKDIKTI) in SINTA 4 No. 36/E/KPT/2019 since 13 December 2019.
Articles 221 Documents
Pengembangan Sistem Visual-To-Speech Berbasis Yolov11, Qwen2.5, dan MobileNetV4 Berbahasa Indonesia Untuk Penyandang Tunanetra Kevin Handoyo; Yosefina Finsensia Riti
Jurnal Pseudocode Vol 13 No 2 (2026): Volume 13 Nomor 2 September 2026
Publisher : UNIB Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33369/pseudocode.13.2.207-216

Abstract

Visually impaired individuals face significant barriers in perceiving their surroundings and conducting cash transaction independently due to their visual limitations. This study develops an android-based assistive system that integrates three capabilities into single pipeline: object detection, contextual visual description in Indonesia, and Indonesia Rupiah Identification, with text-to-speech output, YOLOv11n and YOLOv12n were compared for object detection, Qwen2.5-VL under zero-shot prompting was compared against fine-tuned BLIP-2 for visual description, and MobileNetV4 with a CBAM attention module was fine-tuned for identifying eleven Rupiah note and coin types (seven banknotes and four coins). Evaluations on 5,000 MS-COCO validation images showed YOLOv11n achieving mAP0.5 of 0.5230 with an inference speed of 70.1 FPS, roughly 2.5 times faster than YOLOv12n despite the latter’s marginally higher accuracy, making YOLOv11n the preferred choice for a responsive assistive pipeline. For visual description, Qwen2.5-VL with YOLO detected object labels as prompting hints reached a BLEU-1 score of 0.3720 and CIDEr of 0.1621, outperforming fine-tuned BLIP-2, whose OPT-2.7B backbone and limited 32-token Q-Former query capacity constrained its ability to generate fluent Indonesian captions. The two-stage MobileNetV4+CBAM currency identification module achieved a localizer mAP0.5 of 0.9891 and a classifier overall accuracy of 0.9879 with a macro F1-score of 0.9903, with classification errors concentrated almost entirely among low-value coins rather than between coins and banknotes. These results demonstrate that the Rupiah currency detection module can run entirely on-device without requiring an internet connection, while the object detection and visual description components rely on server-side processing. This hybrid architecture balances low-latency, offline-capable currency recognition with the computational demands of more complex vision-language tasks, offering a practical assistive solution for the visually impaired. Keywords: assistive technology computer vision image captioning object detection Rupiah currency identification