Garuda - Garba Rujukan Digital

Development of Object Detection System on Non-Helmed Riders Using YOLOv8 Prihasto, Bima; Fadhliana, Nisa Rizqiya; Hariyani, Agustina; Alwafi, Fauzan M.; Askarin, Tsaqila B.
Jurnal Pendidikan Multimedia (Edsence) Volume 5 No 2 (December 2023)
Publisher : Universitas Pendidikan Indonesia (UPI)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.17509/edsence.v5i2.65910

Motorcycle accidents are a severe problem, with the number of incidents reaching 66,602 in 2023. Helmets as head protection are mandatory, but awareness of their use is still low. This research utilises Deep Learning, specifically YOLOv8, to detect helmet use violations among motorbike riders. The research results show high accuracy with a Precision of 89.5%, Recall at 78.4%, and mAP50 at 85.7%. YOLOv8 effectively detects violations and provides a solid basis for increasing motorist awareness. Through this innovative approach, it is hoped that a safer driving culture and collective awareness of responsibility in traffic safety will be created.

Comparative Analysis of K-Means and K-Medoids Clustering Methods on Weather Data of Denpasar City Prihasto, Bima; Darmansyah, Darmansyah; Yuda, Dastin Pranata; Alwafi, Fauzan Maftuh; Ekawati, Herliana Nur; Sari, Yustika Perwita
Jurnal Pendidikan Multimedia (Edsence) Volume 5 No 2 (December 2023)
Publisher : Universitas Pendidikan Indonesia (UPI)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.17509/edsence.v5i2.65925

By applying data mining methods, particularly clustering techniques, the weather data of Denpasar city can be grouped based on similar characteristics. This provides deep insight into weather patterns, useful for more optimised travel planning. This research positively impacts tourism, helping stakeholders understand weather patterns in more detail. Furthermore, in-depth knowledge of weather conditions improves preparedness for potential global climate change. The clustering results can be visualised in a three-dimensional cartesian diagram, providing a clear picture of the various weather conditions using attributes such as temperature, precipitation, and humidity. Through Kaggle's Denpasar Weather Data dataset, with 264,924 data and 32 columns, this study illustrates that cloudy weather dominates in the K-Means and K-Medoids clustering process on rain data within one hour. At three hours, K-Means shows the dominance of cloudy weather and the possibility of rain, while K-means dominates all clusters. At six hours, K-Means dominate in sunny and rainy weather, while K-Medoids dominate evenly in all clusters.

Expert System for Diagnosing Plant-disturbing Organisms on Rice Plants Using the Euclidean Probability Method and Bayes Theorem with Forward Chaining Inference Technique Azhar, Nur Fajri; Prihasto, Bima; Nadhira Rizqana Nur Salsabila
SPECTA Journal of Technology Vol. 8 No. 3 (2024): SPECTA Journal of Technology
Publisher : LPPM ITK

Show Abstract | Download Original | Original Source | Check in Google Scholar

Rice is a basic human need that needs to be fulfilled continuously, especially in Indonesia. However, rice production decreased by 2.05% in 2023; the decline was influenced by the lack of rice fields and crop failure due to attacks by plant-disturbing organisms such as Blast, Brown Spot, and even Ricefield Rats. Therefore, expert system technology is useful to help create opportunities for progress in the agricultural sector in overcoming the decline in production. This research utilizes the best method between Euclidean Probability, Bayes` Theorem, and a combination of both in diagnosing plant-disturbing organisms in rice plants. The expert system works by analyzing the symptoms and characteristics of the plants using weight values obtained from the Analytical Hierarchy Process, comparing them with a database of known plant-disturbing organisms, and providing accurate diagnoses and management recommendations. The objectives are to determine which method provides the most accurate diagnosis and to explore how these methods can support sustainable agriculture. The combination of Bayes' theorem with Euclidean methods and Bayes' theorem alone achieved an agreement of 8 out of 10 cases with expert diagnoses. In comparison, the Euclidean method alone achieved an agreement of 9 out of 10 cases. The results demonstrate that the Euclidean Probability method offers a more accurate diagnosis, aligning with expert diagnoses in 9 of the 10 case studies, thus supporting its application in sustainable agricultural practices.

Advancing Voice Anti-Spoofing Systems: Self-Supervised Learning and Indonesian Dataset Integration for Enhanced Generalization Prihasto, Bima; Nur Farid, Mifta; Al Khairy, Rafid
Brilliance: Research of Artificial Intelligence Vol. 4 No. 2 (2024): Brilliance: Research of Artificial Intelligence, Article Research November 2024
Publisher : Yayasan Cita Cendekiawan Al Khwarizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/brilliance.v4i2.5182

This study examines how self-supervised learning and a novel Indonesian language dataset enhance anti-spoofing systems. Results show improved model performance, with a lower Equal Error Rate (EER) during training, indicating effective learning from diverse audio samples. Using weighted cross-entropy analysis highlights the model's robustness in minimizing training errors. Comparisons with baseline models using English data reveal the proposed approach's superiority, achieving a significantly lower EER due to the incorporation of language-specific data. The unique phonetic features of Indonesian languages provide valuable training material, boosting the system's defence against spoofing attacks. The dataset improves generalization across dialects and recording conditions by including diverse speech samples. This integration enhances the anti-spoofing systems' adaptability, which is vital for real-world applications where recording variability affects performance. The experimental setup used a balanced dataset of genuine and spoofed utterances from male and female speakers, ensuring high-quality input. The training configuration splits the dataset into training, development, and testing sets on a high-performance computing setup. Results showed the proposed model achieved an EER of 0.33, compared to 7.65 for the traditional sinc-layer model and 0.82 for the wav2vec 2.0 model with English data. Overall, this research advances anti-spoofing solutions and emphasizes the need for diverse datasets and advanced learning approaches to improve automatic speaker verification systems in practical applications. The incorporation of the Indonesian dataset is vital for addressing linguistic diversity challenges in biometric security, paving the way for future advancements in this area.

Advancing Voice Anti-Spoofing Systems: Self-Supervised Learning and Indonesian Dataset Integration for Enhanced Generalization Prihasto, Bima; Nur Farid, Mifta; Al Khairy, Rafid
Brilliance: Research of Artificial Intelligence Vol. 4 No. 2 (2024): Brilliance: Research of Artificial Intelligence, Article Research November 2024
Publisher : Yayasan Cita Cendekiawan Al Khwarizmi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/brilliance.v4i2.5182

This study examines how self-supervised learning and a novel Indonesian language dataset enhance anti-spoofing systems. Results show improved model performance, with a lower Equal Error Rate (EER) during training, indicating effective learning from diverse audio samples. Using weighted cross-entropy analysis highlights the model's robustness in minimizing training errors. Comparisons with baseline models using English data reveal the proposed approach's superiority, achieving a significantly lower EER due to the incorporation of language-specific data. The unique phonetic features of Indonesian languages provide valuable training material, boosting the system's defence against spoofing attacks. The dataset improves generalization across dialects and recording conditions by including diverse speech samples. This integration enhances the anti-spoofing systems' adaptability, which is vital for real-world applications where recording variability affects performance. The experimental setup used a balanced dataset of genuine and spoofed utterances from male and female speakers, ensuring high-quality input. The training configuration splits the dataset into training, development, and testing sets on a high-performance computing setup. Results showed the proposed model achieved an EER of 0.33, compared to 7.65 for the traditional sinc-layer model and 0.82 for the wav2vec 2.0 model with English data. Overall, this research advances anti-spoofing solutions and emphasizes the need for diverse datasets and advanced learning approaches to improve automatic speaker verification systems in practical applications. The incorporation of the Indonesian dataset is vital for addressing linguistic diversity challenges in biometric security, paving the way for future advancements in this area.

PURE PALM OIL SPRAY CHARACTERISTICS OF DIESEL FUEL INJECTOR Mamola, Randi; Nurcholik, Samsu Dlukha; Prahmana, Rico Aditia; Juangsa, Firman Bagja; Pratama, Raditya Hendra; Prihasto, Bima; Suardi, Suardi; Sa’adiyah, Devy Setiorini
Jurnal Rekayasa Mesin Vol. 16 No. 1 (2025)
Publisher : Jurusan Teknik Mesin, Fakultas Teknik, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.21776/jrm.v16i1.1981

Diesel engines are widely used in many sectors due to their advantages of high energy density. To reduce emissions in diesel engines, a step that can be taken is to use renewable biofuel such as pure palm oil, which has a high viscosity and is difficult to get a fine atomization process in fuel spraying. However, a detailed analysis of the spray characteristics using pure palm oil fuel on conventional diesel engines has yet to be available. In this research, high-speed imaging was conducted to investigate the spray characteristics of pure palm oil compared to diesel fuel. The result shows that (1) at the start of the injection stage, both pure palm oil and diesel fuel have similar shapes, (2) the spray angle of pure palm oil is smaller due to orifice wall cavitation not occurring in the orifice, (3) and at the end of the injection stage, pure palm oil has larger droplets, which is also more difficult to atomize.

Pemanfaat Kecerdasan Buatan dalam Pengembangan Pembelajaran di Daerah Penyangga Ibukota Negara (IKN) Indira Anggriani; Riyan Benny Sukmara; Bima Prihasto; Nur Fajri Azhar
Ininnawa : Jurnal Pengabdian Masyarakat Vol. 1 No. 2 (2023): Volume 01 Nomor 02 (Oktober 2023)
Publisher : Program Studi Manajemen FEB UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26858/ininnawa.v1i2.567

Di era digital, teknologi menjadi bagian yang tak terpisahkan dari kehidupan sehari – hari, termasuk dalam dunia pendidikan. Oleh karena itu, salah satu upaya yang dapat dilakukan yaitu mengelaborasi pemanfaatan teknologi dalam pembelajaran di sekolah. Salah satu kendala yang dihadapi adalah kurangnya waktu dan sumber daya dalam membuat soal ujian dan tes. Untuk mengatasi kendala tersebut, diperlukan inovasi dalam pembelajaran di era digital. Salah satu inovasi yang dapat dilakukan adalah dengan penggunaan teknologi OpenAI API pada pembuatan soal otomatis, dapat memperbaiki masalah yang masih terbatasnya SDM guru serta pada pembuatan soal ujian yang masih dilakukan secara manual di sekolah daerah penyangga IKN. Pada pelatihan ini, para guru diajarkan dalam penggunaan teknologi OpenAI API untuk mendukung pembelajaran di sekolah, khususnya di daerah penyangga Ibukota Negara (IKN).

Pembelajaran Berbasis Teknologi dalam Meningkatkan Kemampuan dan Minat Belajar Anak Desa Sei Wain Prihasto, Bima; Winarni; Kamil, Muhammad Insan; Febriandini, Nanda Clariza; Pratama, Abdul Rizal; Himawan, Kevin; Akbar, Muhammad; Syakbani, Ahmad Rusdianto Andarina; Edia, Naflah Shafa
Ininnawa : Jurnal Pengabdian Masyarakat Vol. 2 No. 2 (2024): Volume 02 Nomor 02 (Oktober 2024)
Publisher : Program Studi Manajemen FEB UNM

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26858/ininnawa.v2i2.4448

Kegiatan pengabdian masyarakat ini bertujuan untuk meningkatkan pemahaman dan minat belajar anak-anak di RT 36 Kelurahan Karang Joang, Kecamatan Balikpapan Utara, terhadap bahasa Inggris, matematika, dan pemrograman. Dengan pendekatan interaktif dan berbasis gamifikasi, melalui penggunaan aplikasi Duolingo, CodeCombat, dan Kahoot, anak-anak diajak untuk belajar dengan cara yang menyenangkan dan menantang. Evaluasi program menunjukkan peningkatan yang signifikan dalam pemahaman matematika dan bahasa Inggris, serta dalam membangun rasa percaya diri dalam menggunakan bahasa Inggris. Untuk penyempurnaan program di masa mendatang, disarankan untuk melakukan evaluasi yang lebih terperinci, melibatkan lebih banyak pihak, dan memperluas jangkauan program.

Evolution of the Intellectual Property Information System at the Kalimantan Institute of Technology Using the Waterfall Method and Design Thinking Rajab, Nur Ali; Marianta, Arwin; Lestari , Ika; Azhar, Nur Fajri; Prihasto, Bima
Sebatik Vol. 29 No. 2 (2025): December 2025
Publisher : STMIK Widya Cipta Dharma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46984/sebatik.v29i2.2681

Intellectual Property (IP) Management at the Kalimantan Institute of Technology (ITK) was previously hindered by manual processes using Google Forms and Excel that were inefficient and prone to errors, and by limited information systems. This research aims to evolve the ITK Intellectual Property Information System (SIM KI) to enhance data management efficiency at the backend and improve functionality and user experience at the frontend. The development methodology uses a structured Waterfall approach. The backend is developed with Node.js (Express) and PostgreSQL, while the frontend uses React JS with interface design based on Design Thinking. System verification is conducted comprehensively through Black Box Testing, White Box Testing, User Acceptance Testing (UAT), and the System Usability Scale (SUS). The evolution results show successful system implementation that now supports four types of IP (Copyright, Patent, Trademark, Industrial Design). Backend testing through unit testing validates the reliability of internal logic, while frontend testing demonstrates functional success (Black Box), good usability (SUS score 77.89), and significant user acceptance improvement (UAT score increased from 68% to 76%). The evolution of SIM KI successfully resulted in a more efficient, functional, and well-received digital platform, which implies increased effectiveness of IP management in the academic environment of ITK.

Automatic identification of herbal medicines using deep learning on leaf images Anita Ahmad Kasim; Lukman Nadjamudiin; Muhammad Bakri; Chairunnisa Ardiansyah Lamasitudju; Harry Tanni Pagiu; Puguh Budi Prakoso; Anindita Septiarini; Bima Prihasto
International Journal of Advances in Intelligent Informatics Vol 12, No 2 (2026): May 2026
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26555/ijain.v12i2.2024

Indonesia has a high diversity of medicinal plants that are widely used in traditional healthcare practices. Identification of medicinal plants is commonly based on leaf morphology; however, similarities in leaf shape, texture, and color often cause misidentification, particularly among non-experts. This limitation highlights the need for an automated and reliable identification approach. The primary objective of this study is to develop and evaluate a deep learning–based system for the automatic identification of medicinal plants using leaf images, with a specific focus on comparing the performance and efficiency of MobileNetV2 and ResNet50V2 architectures. The research design adopts an experimental approach using an internally collected dataset of medicinal plant leaf images representing multiple plant classes. The dataset is divided into training and testing sets to evaluate model generalization. The methodology involves image preprocessing steps, including resizing, normalization, and data augmentation, followed by the application of transfer learning using MobileNetV2 and ResNet50V2 as feature extractors. Both models are trained under the same experimental settings and evaluated using standard classification metrics, including accuracy, precision, recall, F1-score, and confusion matrix analysis. The main outcomes and results indicate that both deep learning models achieve high classification performance. MobileNetV2 achieves an accuracy of 98.77%, precision of 98.84%, recall of 98.77%, and F1-score of 98.77%, while ResNet50V2 achieves an accuracy of 97.53%, precision of 97.87%, recall of 97.53%, and F1-score of 97.58%. The results demonstrate that MobileNetV2 provides slightly superior performance with lower computational complexity. In conclusion, lightweight deep learning architectures such as MobileNetV2 are effective and efficient for medicinal plant leaf identification and are suitable for implementation in mobile or resource-constrained environments.

Title

Found 13 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 13 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 13 Documents
Search