Claim Missing Document
Check
Articles

Found 20 Documents
Search

Revealing the Characteristics of Balinese Dance Maestros by Analyzing Silhouette Sequence Patterns Using Bag of Visual Movement with HoG and SIFT Features Made Windu Antara Kesiman; I Made Dendi Maysanjaya; I Made Ardwi Pradnyana; I Made Gede Sunarya; Putu Hendra Suputra
Journal of ICT Research and Applications Vol. 15 No. 1 (2021)
Publisher : LPPM ITB

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5614/itbj.ict.res.appl.2021.15.1.6

Abstract

The aim of this research was to reveal and explore the characteristics of Balinese dance maestros by analyzing silhouette sequence patterns of Balinese dance movements. A method and complete scheme for the extraction and construction of silhouette features of Balinese dance movements are proposed to enable performing quantitative analysis of Balinese dance movement patterns. Two different feature extraction methods, namely the Histogram of Gradient (HoG) feature and the Scale Invariant Features Transform (SIFT) descriptor, were used to build the final feature, called the Bag of Visual Movement (BoVM) feature. This research also makes a technical contribution with the proposal of quantifying measures to analyze the movement patterns of Balinese dances and to create the profile and characteristics of dance maestros/creators. Eight Balinese dances from three different Balinese dance maestros were analyzed in this work. Based on the experimental results, the proposed method was able to visually detect and extract patterns from silhouette sequences of Balinese dance movements. Quantitatively, the pattern measures for profiling of Balinese dances and maestros revealed a number of significant characteristics of different dances and different maestros.
Intisari Kerangka Sistem Berbasis Aturan Menggunakan Certainty Factor Dengan Runut Maju Dan Runut Mundur Putu Hendra Suputra
Jurnal Nasional Pendidikan Teknik Informatika : JANAPATI Vol. 1 No. 1 (2012)
Publisher : Prodi Pendidikan Teknik Informatika Universitas Pendidikan Ganesha

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23887/janapati.v1i1.9753

Abstract

Salah satu cara dalam membangun suatu sistem pakar adalah dengan menggunakankerangka (shell) yang membantu membangun knowledge base dan aplikasinya. Kerangkasistem pakar membantu pengembang sistem pakar sebagai fasilitas antarmuka pengguna dalam mengelola input dan output dari data. Kerangka sistem pakar juga mampu memanipulasi informasi dan basis pengetahuan yang dimiliki, untuk menghasilkan sebuah konklusi. Di samping fasilitas konstruksi pengetahuan yang mudah dan memadai, pengembangan sistem pakar dengan menggunakan kerangka (shell) sistem pakar juga perlu memperhatikan fasilitas penjelasan yang ramah penguna dan mampu menjelaskan bagaimana suatu kesimpulan diperoleh.MyExSys adalah sebuah kerangka sistem berbasis aturan yang dikembangkan untuk membangun sebuah sistem pakar berbasis aturan (pengetahuan) yang memberi keleluasaan kepada pengguna untuk mengelola pengetahuan. Proses penalaran yang didukung adalah runut maju dan runut mundur serta jejaring kemungkinan Fasilitas penjelasan yang disediakan meliputi rangkuman proses penalaran yang berbasis teks, serta jejaring inferensi dalam bentuk grafik AND-OR Tree. MyExSys menggunakan certainty factor secara naif dalam merepresentasikan penaran atas ketidakpastian.
Automatic 3D Cranial Landmark Positioning based on Surface Curvature Feature using Machine Learning Putu Hendra Suputra; Anggraini Dwi Sensusiati; Myrtati Dyah Artaria; Gijsbertus Jacob Verkerke; Eko Mulyanto Yuniarno; I Ketut Eddy Purnama
Knowledge Engineering and Data Science Vol 5, No 1 (2022)
Publisher : Universitas Negeri Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.17977/um018v5i12022p27-40

Abstract

Cranial anthropometric reference points (landmarks) play an important role in craniofacial reconstruction and identification. Knowledge to detect the position of landmarks is critical. This work aims to locate landmarks automatically. Landmarks positioning using Surface Curvature Feature (SCF) is inspired by conventional methods of finding landmarks based on morphometrical features. Each cranial landmark has a unique shape. With the appropriate 3D descriptors, the computer can draw associations between shapes and landmarks using machine learning. The challenge in classification and detection in three-dimensional space is to determine the model and data representation. Using three-dimensional raw data in machine learning is a serious volumetric issue. This work uses the Surface Curvature Feature as a three-dimensional descriptor. It extracts the local surface curvature shape into a projection sequential value (depth). A machine learning method is developed to determine the position of landmarks based on local surface shape characteristics. Classification is carried out from the top-n prediction probabilities for each landmark class, from a set of predictions, then filtered to get pinpoint accuracy. The landmark prediction points are hypothetically clustered in a particular area, so a cluster-based filter is appropriate to isolate them. The learning model successfully detected the landmarks, with the average distance between the prediction points and the ground truth being 0.0326 normalized units. The cluster-based filter is implemented to increase accuracy compared to the ground truth. Thus, SCF is suitable as a 3D descriptor of cranial landmarks.
Deteksi Jatuh Lansia Berbasis Landmark Sendi Pada Model LSTM Pratama Putra, Gede Bakti; Suputra, Putu Hendra; Marti, Ni Wayan; Sugiantari, Kadek Feny
Jurnal Informatika Polinema Vol. 11 No. 3 (2025): Vol. 11 No. 3 (2025)
Publisher : UPT P2M State Polytechnic of Malang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33795/jip.v11i3.6823

Abstract

Fenomena lansia terlantar di Indonesia terus meningkat, dengan 2,4 juta dari 25 juta lansia terlantar pada tahun 2019 menurut data CAS Universitas Indonesia. Dengan populasi lansia yang terus bertambah setiap tahun, tantangan dalam perawatan lansia menjadi perhatian penting, khususnya dalam memberikan keamanan dan respons cepat terhadap kondisi darurat. Salah satu kondisi darurat yang sering terjadi adalah jatuh, yang dapat berakibat fatal jika tidak ditangani segera. Penelitian ini bertujuan untuk mengembangkan model deteksi jatuh pada lansia menggunakan algoritma Long Short Term Memory (LSTM) dan fitur landmark sendi. Kondisi jatuh didefinisikan sebagai perubahan pose dari berdiri ke terbaring, yang dipantau secara kontinu menggunakan konsep sliding window. Setiap pose diberikan indeks tertentu yang merepresentasikan kondisi lansia, seperti berdiri, bungkuk, jongkok, duduk, terbaring kiri, terbaring kanan, terbaring ke atas, dan terbaring ke bawah. Model LSTM digunakan untuk mengklasifikasikan pose lansia berdasarkan data pose tersebut. Metode penelitian meliputi studi literatur, analisis, pengembangan model, serta evaluasi performa. Penelitian ini menggunakan dataset berupa 240 video dengan 30 frame per video, melibatkan validator dari Panti Sosial X untuk memastikan keakuratan data yang digunakan. Hasil uji coba menunjukkan bahwa model yang dikembangkan mampu mengklasifikasikan pose dan mendeteksi kondisi jatuh dengan akurasi 91%, yang lebih unggul dibandingkan penelitian sebelumnya dengan akurasi model yaitu 86%.
PENDEKATAN MLP DALAM KLASIFIKASI BAHASA ISYARAT: ANALISIS JARAK EUCLIDEAN LANDMARK TANGAN Sugiantari, Kadek Feny; Suputra, Putu Hendra; Dewi, Luh Joni Erawati; Putra, Gede Bakti Pratama
Jurnal Informatika Vol 9, No 2 (2025): JIKA (Jurnal Informatika)
Publisher : University of Muhammadiyah Tangerang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31000/jika.v9i2.13368

Abstract

Perkembangan teknologi Computer Vision dalam Kecerdasan Buatan (AI) mendorong inovasi teknologi yang inklusif dalam komunikasi bagi penyandang disabilitas, seperti tunarungu dan tunawicara. Penelitian ini mengembangkan model klasifikasi bahasa isyarat angka SIBI, khususnya angka 0-9 yang menggunakan jarak Euclidean antar landmark tangan sebagai fitur. Proses penelitian mencakup pengumpulan data, ekstraksi landmark tangan dengan Mediapipe, ekstraksi fitur jarak Euclidean, pelatihan model dengan Multi Layer Perceptron, evaluasi, dan implementasi real-time. Hasil penelitian menunjukkan model berhasil mengklasifikasikan pose angka 0-9 dan non-pose, dengan akurasi 87.17% dan penerapan threshold pada tahap evaluasi serta implementasi real-time untuk memastikan semua input data terklasifikasi dengan tingkat kepercayaan tinggi. Hasil penelitian ini dapat menunjang proses pembelajaran bahasa isyarat SIBI bagi penyandang disabilitas.
Pelatihan Penguatan Kemampuan Berbahasa Jepang Level Chukyu bagi Guru Bahasa Jepang di SMA dan SMK Kabupaten Buleleng Bali Desak Made Sri Mardani; I Kadek Antartika; Putu Hendra Suputra; I Wayan Sadyana
Jurnal Pengabdian UNDIKMA Vol. 4 No. 1 (2023): February
Publisher : LPPM Universitas Pendidikan Mandalika (UNDIKMA)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33394/jpu.v4i1.5713

Abstract

The purpose of this community service is to strengthen the Japanese language skills (Chukyu level) of Japanese High School/Vocational High School Teachers in Buleleng Regency so that their abilities increase and can pass the Teacher Professional Education/PPG. The method used in this activity is the Technical Assistance Model in the form of Training which is carried out by providing training and assistance to improve Japanese language skills. The partners in this activity were 20 (twenty) High School/Vocational High School Japanese language teachers in Buleleng Regency. The evaluation instrument for this activity is a test to determine the teacher's understanding of the material provided, as well as an assessment questionnaire by the teacher to evaluate the implementation of the activity. The results of tests and questionnaires were analyzed descriptively. The results of this activity show that through training and mentoring, the teachers’ skills in reading discourse, knowledge of Japanese culture, and Japanese language proficiency increased. A good score on the test as well as a positive response given by the teacher shows the success of this activity.
A Analisis Tingkat Kepuasan Konsumen Pada Pelayanan PT. AXZ Furniture Di Media Internet Menggunakan Metode VADER dan ARM Laksmi, Ida Ayu; I Made Candiasa; Putu Hendra Suputra
JURNAL RISET KOMPUTER (JURIKOM) Vol. 12 No. 6 (2025): Desember 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/jurikom.v12i6.9322

Abstract

Bali Art Furniture is one of the companies in Bali that exports furniture and home decoration products. During its operation, this company utilises various internet media platform services such as websites, Facebook Marketplace, Instagram, WhatsApp Business, Pinterest, and Google to communicate online with consumers. As the company has grown over time, it has recruited many employees to increase its capacity to provide services. The company has never conducted a systematic evaluation of customer service satisfaction, either internally or externally. Based on this phenomenon, a customer satisfaction analysis was conducted using customer comment data. Customer service satisfaction was evaluated using the VADER and ARM methods as a basis for comparing the effectiveness of these methods. Based on the analysis of the two methods, the VADER method produced an accuracy of 34%, while the ARM method produced an accuracy of 64%. The evaluation results using the confusion matrix of the VADER model showed that positive comments were more recognisable by the system than negative and neutral comments, as seen from the positive recall value of 0.90, which was greater than the negative and neutral recall values. Meanwhile, the evaluation results using the ARM method showed that neutral comments were more recognisable by the system than positive and negative comments, as seen from the neutral recall value of 0.88, which was greater than the positive and negative recall values. Thus, the highest accuracy results in the ARM model became the guideline in making recommendation results.
SMART LOCK SYSTEM BERBASIS IOT MENGGUNAKAN ESP32 UNTUK KEAMANAN AKSES PUSAT DATA (STUDI KASUS: UPA TIK UNDIKSHA) Aryadi, Made Waradiana; Arthana, I Ketut Resika; Suputra, Putu Hendra
Jurnal Informatika dan Teknik Elektro Terapan Vol. 14 No. 1 (2026)
Publisher : Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23960/jitet.v14i1.8826

Abstract

Keamanan fisik pusat data berperan penting dalam menjamin keberlangsungan layanan teknologi informasi dari ancaman akses tidak sah dan tindakan kriminal. Pusat data UPA TIK Universitas Pendidikan Ganesha telah menerapkan sistem pengamanan dasar, namun masih menghadapi keterbatasan pada aspek pemantauan real-time, notifikasi otomatis, dan pengelolaan akses pengguna. Penelitian ini bertujuan merancang dan mengembangkan smart lock system berbasis Internet of Things menggunakan ESP32 guna meningkatkan keamanan akses fisik pusat data. Sistem yang dikembangkan mengintegrasikan autentikasi RFID dan sidik jari, komunikasi antar perangkat melalui protokol MQTT, sensor magnetic door switch untuk mendeteksi akses ilegal, serta notifikasi real-time melalui bot Telegram. Sistem juga dilengkapi fitur pendaftaran pengguna secara dinamis tanpa perlu pemrograman ulang perangkat. Metode penelitian yang digunakan adalah Research and Development. Hasil pengujian menunjukkan seluruh fungsi sistem berjalan sesuai spesifikasi dengan rata-rata latency sekitar 4 detik dan tingkat keberhasilan autentikasi 100%. RFID menawarkan respons yang cepat dan stabil namun memiliki risiko duplikasi dan kehilangan kartu, sedangkan sidik jari memberikan keamanan lebih tinggi meskipun latency-nya bervariasi. Secara keseluruhan, sistem ini mampu meningkatkan keamanan fisik pusat data secara terintegrasi dan responsif.
Analisis Sentimen Penggunaan Cekat.AI dalam Menggantikan Customer Service Menggunakan Logistic Regression dan TF-IDF Ardyaputra, Gede Yudha; Pascima, Ida Bagus Nyoman; Suputra, Putu Hendra
Journal of Computer System and Informatics (JoSYC) Vol 7 No 2 (2026): February 2026
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/josyc.v7i2.9292

Abstract

The rapid development of artificial intelligence has significantly transformed customer service systems, particularly through the use of chatbots to replace human customer service agents. Cekat.AI is one of Indonesia’s local AI-based chatbot innovations that has been increasingly adopted by companies. However, its implementation has generated diverse public reactions on social media platforms, especially X and TikTok. The main problem addressed in this study is how users perceive and respond sentimentally to the use of Cekat.AI as a replacement for human customer service, as well as the underlying factors influencing these ssentiments This study aims to analyze public sentiment using the Logistic Regression method with Term Frequency–Inverse Document Frequency (TF-IDF) feature extraction on social media comments from X and TikTok. To address class imbalance in sentiment data, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. The results indicate that negative sentiment dominates at 52.5%, followed by positive sentiment at 35.1% and neutral sentiment at 12.4%. The implementation of SMOTE significantly improved the Recall of the neutral class from 18.6% to 64.1%, with a cross-validation accuracy of 79.98%. Topic modeling further reveals that negative sentiment is primarily driven by automation anxiety and concerns over job displacement. These findings suggest that the main challenge in adopting Cekat.AI lies in social acceptance rather than technical performance. This study provides a dual contribution, namely technically proving the effectiveness of SMOTE in handling extreme imbalance in Indonesian text data, and practically revealing that public resistance to local AI is rooted in job displacement anxiety, not merely technical service aspects.
Rancang Sistem Deteksi Dini Terjadinya Kebakaran Menggunakan Logika Fuzzy pada Area Parkiran I Gusti Putu Kresna Dana; Putu Hendra Suputra
Jurnal Nasional Teknologi Informasi dan Aplikasinya Vol. 3 No. 3 (2025): JNATIA Vol. 3, No. 3, Mei 2025
Publisher : Informatics Department, Faculty of Mathematics and Natural Sciences, Udayana University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24843/JNATIA.2025.v03.i03.p14

Abstract

This study discusses the development of a fire monitoring system in a semi-outdoor multi-storey car park by utilizing Internet of Things (IoT) technology. The experimental research method is used with the design science research method (DSRM) stage. The system uses MQ-2, DHT11, and YG-1006 sensors to detect smoke, temperature increases, and flames. Data from the sensors is processed using fuzzy logic to determine the level of danger by the Arduino Uno microcontroller and notifications are forwarded via the LCD display to security officers. The test results show that the system can produce digital values that represent the level of danger based on environmental conditions. The conclusion states that this system has the potential to improve responsiveness and reliability in dealing with fire risks in a semi-outdoor multi-storey car park environment.