Claim Missing Document
Check
Articles

Found 8 Documents
Search

Deep Learning Verifikasi Kemiripan Wajah Dengan Arsitektur Jaringan Siamese Kartarina Kartarina; Hairul Imam
Jurnal Bumigora Information Technology (BITe) Vol 1 No 2 (2019)
Publisher : Prodi Ilmu Komputer Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (947.236 KB) | DOI: 10.30812/bite.v1i2.605

Abstract

Abstrak Verifikasi wajah adalah masalah yang cukup populer dalam bidang computer vision. Banyak pendekatan yang telah dilakukan untuk menyelesaikan masalah tersebut baik menggunakan model matematika murni dengan mempelajari pola geometri pada wajah secara manual maupun cara otomatis menggunakan pendekatan pembelajaran mesin. Penelitian ini mencoba memecahkan masalah tersebut dengan pendekatan deep learning, dimana model dilatih menggunakan triplet loss yang didefinisikan pada paper FaceNet. Rancangan model yang digunakan adalah Siamese dengan menerapkan ResNet-50 yang telah dimodifikasi untuk mempelajari fitur yang ada pada gambar sehingga mampu mereduksi dimensi gambar yang tinggi menjadi vektor baris yang rendah disebut sebagai embedding. Setelah model berhasil mempelajari embedding yang baik pada gambar maka masalah verifikasi wajah bisa diselesaikan dengan membandingkan jarak embedding antar gambar dimana jarak yang dekat dapat diartikan sebagai wajah yang mirip (genuine) dan jarak yang jauh dapat diartikan sebagai wajah yang berbeda (impostor). Pada penelitian ini, model berhasil dilatih pada Dataset VGG Face v2 (Visual Geometry Group) dengan nilai akurasi 92% pada Dataset LFW (Labeled Faces in the Wild) sebagai data testing dan mendapatkan nilai AUC (Area Under the Curve) 97%. Nilai AUC yang tinggi dapat diartikan bahwa model dapat memverifikasi dengan baik gambar wajah orang yang sama sebagai genuine dan gambar wajah orang yang berbeda sebagai impostor. Kata Kunci: Siamese, Triplet Loss, Verifikasi Wajah, Face Embedding, Dimensionality Reduction. Abstract Face verification is a quiet popular problem in computer vision. Various approach has been applied to solve this problem from using pure mathematical model by manually defining Face geometric pattern into automatic way by using machine learning. This research try to solve this problem using deep learning approach. The model will be trained using Triplet Loss as defined in the Face Net paper. The model architecture that will be used is Siamese by applying modified ResNet-50 as the body of the Network, the Network will be trained as how to reduce high dimension image into a low dimension row vector, reduced image dimension into a low row feature vector also called embedding. If model successfully trained to produce a good embedding quality of an image then Face verification problem can be seen as Pythagorean problem where the distance of two pair of images can be calculated using euclidean distance those the distance can be seen as the similarity value which by applying some threshold value we can determine if those pair of images is genuine (similar) or not (impostor). This Research, model successfully trained on VGG Face v2 (Visual Geometry Group) Dataset by getting 92% accuracy on LFW (Labelled Face in the Wild) as testing Dataset. Also the AUC (Area Under The Curve) score is reacing 97%, high AUC score can be interpreted that the model is successfully verify similar person as genuine and different person as impostor. Keywords: Siamese, Triplet Loss, Face Verification, Face Embedding, Dimensionality Reduction.
Analysis of Document Clustering based on Cosine Similarity and K-Main Algorithms Bambang Krismono Triwijoyo; Kartarina Kartarina
Journal of Information System and Informatics Vol 1 No 2 (2019): Journal of Information Systems and Informatics
Publisher : Universitas Bina Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33557/journalisi.v1i2.18

Abstract

Clustering is a useful technique that organizes a large number of non-sequential text documents into a small number of clusters that are meaningful and coherent. Effective and efficient organization of documents is needed, making it easy for intuitive and informative tracking mechanisms. In this paper, we proposed clustering documents using cosine similarity and k-main. The experimental results show that based on the experimental results the accuracy of our method is 84.3%.
SISTEM INFORMASI PEMASARAN PRODUK DESA BERBASIS WEB Kartarina Kartarina; Pahrul Irfan; Cristofer Satria
ILKOM Jurnal Ilmiah Vol 11, No 3 (2019)
Publisher : Teknik Informatika Fakultas Ilmu Komputer Univeristas Muslim Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33096/ilkom.v11i3.471.214-221

Abstract

Batu Mekar Village is one of the villages in West Lombok Regency in Lingsar District. Batu Mekar village has many products in the form of handicrafts such as bags, lamp decorations made from ketak or agricultural and plantation products such as coffee, palm sugar and others. The problem that exists in Batu Mekar Village is the low sales price of handicraft products because the product is only left at a souvenir shop. The decrease in the number of sales was not due to a decrease in the number of tourists who ended up on the island of Lombok due to the earthquake. In addition, the limited promotion media. This research was developed using the Rapid Application Development method. This research resulted in an online sales information system as a medium for handicraft sales, so that the public could directly sell their products without going through intermediaries as well as promotional media so that handicraft products can be sold directly and can be recognized by the wider community. The developed application can be used as a promotional medium as well as a means of selling Batu Mekar village handicraft products.
Sosialisasi dan Pengenalan Computational Thinking kepada Guru pada Program Gerakan Pandai oleh Bebras Biro Universitas Bumigora Kartarina Kartarina; Miftahul Madani; Diah Supatmiwati; Regina Aprilia Riberu; Indah Puji Lestari
ADMA : Jurnal Pengabdian dan Pemberdayaan Masyarakat Vol 2 No 1 (2021): Juli 2021
Publisher : LPPM Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/adma.v2i1.1271

Abstract

Currently school teachers, especially in Mataram City, West Lombok Regency, Central Lombok and East Lombok are not familiar with learning with the concept of computational thinking (Computational Thinking) so they cannot teach their students how to think computationally as an approach to solving existing problems. Considering one of the demands of the industrial revolution 4.0, where problem solving skills are one of the abilities that students must have. In this case, these abilities need to be taught by teachers at school. Therefore, this problem must be solved immediately by increasing the ability of teachers in learning computational thinking so that teachers can apply computational thinking learning methods to their students. From the problems listed, it is necessary to approach how to train teachers to teach computationally thinking to their students. In Lombok, West Nusa Tenggara, to apply Computational Thinking (CT) in formulating problems and revealing solutions, namely through socialization and training and mentoring of free computational thinking materials to teachers in schools in Lombok, NTB which was held in the form of CT Bebras socialization activities, which is expected to help introduce and apply Computational Thinking (CT) material as a creative learning method in schools in NTB.
Analisis Dependency Injection dan Model-View-Presenter Pada Aplikasi Berbasis Android Kartarina Kartarina; Apriliansyah Apriliansyah
Progresif: Jurnal Ilmiah Komputer Vol 18, No 1: Februari 2022
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/progresif.v18i1.781

Abstract

Abstrak. Coupling merupakan salah satu indikator kualitas perancangan perangkat lunak terstruktur dan berorientasi objek. Coupling yang ketat akan sulit dirawat sehingga langkah awal mengendalikan Coupling yang baik adalah dengan menerapkan Design Pattern. Eksperimen penerapan Design Pattern membuktikan bahwa design pattern tidak menjamin pengkodean dapat lepas dari Design Smell. Masalah Coupling pada objek yang berkaitan dengan sumber data adalah ketika objek membutuhkan referensi objek Android Context. Android mengumumkan penggunaan Hilt sebagai standar dalam otomatisasi Dependency Injection. Penerapan MVP dapat membantu pengiriman referensi objek Android Context ke Model. Penelitian ini bertujuan menganalisis lebih lanjut untuk mengetahui hasil komparasi metrik kopling objek dari penerapan Dependency Injection secara manual dan otomatisasi pada aplikasi Android. Penelitian ini menghasilkan, otomatisasi penambahan jumlah kopling kelas pada View dan mengurangi kopling kelas pada Presenter. Selain itu, manfaat dari penerapan Dependency Injection secara otomatisasi mengurangi tanggung jawab Class pada Presenter dan menghilangkan Design Smell atau desain yang kompleks, walau begitu, penerapan Dependency Injection secara otomatisasi mempengaruhi pola MVP yang menggunakan Dependency Injection secara manual.Kata kunci: Coupling; Design Pattern; Dependency Injection; Model-View-Presenter Abstract. Coupling is one indicator of the quality for structured and object-oriented software design. Tight couplings will be difficult to maintain so the first step when controlling a good coupling is to apply a Design Pattern. Experiments on the application of Design Patterns prove that design patterns do not guarantee that coding can be separated from Design Smell. Coupling problem with objects related to data sources is that when the object requires an Android Context object reference. Android announced the use of Hilt as standard in Dependency Injection automation. Implementing MVP can help pass Android Context object references to Models. This aims to analyze further to find out the results of the comparison of object coupling metrics from the manual and automated application of Dependency Injection on Android applications. The results of the research, automatic increasing the number classes of coupling on the View and reduce class coupling on the Presenter. In addition, the benefits of implementing Dependency Injection automatically reduces class responsibility on Presenter and Design Smell or complex designs, however, automated Dependency Injection implementation affects the MVP pattern that uses Dependency Injection manually.Keywords: Coupling; Design Pattern; Dependency Injection; Model-View-Presenter
Rancang Bangun Game Edukasi Bermain Dan Belajar Bahasa Inggris Berbasis Mobile Untuk Anak Usia Dini Melati Rosanensi; Kartarina Kartarina; Wisnu Yunanda
Jutisi : Jurnal Ilmiah Teknik Informatika dan Sistem Informasi Vol 11, No 1: April 2022 (Artikel/Judul Accepted)
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/jutisi.v11i1.764

Abstract

Abstrak. Pembelajaran Bahasa Inggris bagi anak usia dini pada PAUD Semai Harapan Bangsa, selama ini dilakukan secara langsung di kelas dengan metode bermain tebak lagu dan praktik percakapan langsung. Namun dengan adanya kebijakan pembatasan akibat Pandemi Covid-19, metode ini tidak dapat dilakukan, sehingga perlu metode pembelajaran alternatif agar pembelajaran dapat berlangsung secara efektif. Makalah ini mengusulkan model game edukasi bermain dan belajar Bahasa Inggris berbasis mobile bagi anak usia dini untuk mengefektifkan proses belajar di masa Pandemi Covid-19. Aplikasi dikembangkan dengan metode pengembangan perangkat lunak multimedia versi Luther Sutopo. Aplikasi dibuat menggunakan software Adobe Ilustrator CS6 untuk mendesai antar muka dan gambar lainnya, sedangkan audio menggunakan fitur Google Voice. MIT App Inventor merupakan platform yang digunakan dalam pembuatan aplikasi. Fitur fungsional dalam aplikasi Game yang diusulkan berupa fitur belajar dan fitur Hiburan berupa Tebak Lagu, Tebak Gambar, dan Ngobrol Santai. Interpretasi hasil uji coba lapangan berada dalam rentang nilai 3,33 – 4,16, yang berarti “Layak” untuk digunakan.Kata kunci: Aplikasi Multimedia; Multimedia Development Life Cycle; Luther Sutopo; Adobe Ilustrator Abstract. Learning English for early childhood at Early childhood education programs “PAUD Semai Harapan Bangsa”, has so far been carried out directly in the classroom using the guessing song method and direct conversation practice. However, with the policy of restrictions due to the Covid-19 pandemic, this method cannot be carried out, so alternative learning methods are needed so that learning can take place effectively. This paper proposes a mobile-based educational game model for playing and learning English for early childhood to streamline the learning process during the Covid-19 pandemic. The application was developed using the Luther Sutopo version of the multimedia software development method. The application is made using Adobe Illustrator CS6 software to design the interface and other images, while the audio uses the Google Voice feature. MIT App Inventor is a platform used in making applications. Functional features in the proposed Game application are learning features and entertainment features in the form of guessing songs, guessing pictures, and chatting casually. The interpretation of the results of field trials is in the range of values from 3.33 to 4.16, which means "Eligible" to be used.Keywords: Multimedia Application; Multimedia Development Life Cycle; Luther Sutopo; Adobe Illustrator
Penerapan Computational Thinking pada Pelajaran Matematika di Madratsah Ibtidaiyah Nurul Islam Sekarbela Mataram Apriani Apriani; Ismarmiaty Ismarmiaty; Dyah Susilowati; Kartarina Kartarina; Wiya Suktiningsih
ADMA : Jurnal Pengabdian dan Pemberdayaan Masyarakat Vol 1 No 2 (2021): ADMA: Jurnal Pengabdian dan Pemberdayaan Masyarakat
Publisher : LPPM Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1183.963 KB) | DOI: 10.30812/adma.v1i2.1017

Abstract

This service activity aims to contribute knowledge to teachers to be able to understand and implement computational thinking in the subjects they are taught. The lack of trained personnel and the lack of understanding in implementing computational thinking gives the Bebras Bureau the opportunity to contribute. This is in line with Mendikbud's desire to implement computational thinking in the children's education curriculum as a provision for more innovative learning to answer the needs of the industrial era 4.0. Computational thinking is the process of thinking in formulating a problem and its solution so that the solution can be represented in a form that can be executed by an information-processing agent. The implementation of the service was carried out on the Mathematics subject teacher at Nurul Islam Mataram Elementary School. The implementation stages consist of planning, preparation, socialization, training, and evaluation. The results of the evaluation showed that most of the teacher participants agreed to apply the results of the training to students and most participants agreed to join the follow-up programs from Bebras. It is hoped that this activity can run continuously and be supported positively by the parties involved.
Pendampingan Proses Pembuatan Soal Berbasis Computational Thinking kepada Guru pada Guru-Guru Tingkat SD dan SMP Kecamatan Sakra, Kabupaten Lombok Timur Ni Ketut Sriwinarti; Apriani Apriani; Diah Supatmawati; Kartarina Kartarina; Ismarmiaty Ismarmiaty
ADMA : Jurnal Pengabdian dan Pemberdayaan Masyarakat Vol 2 No 2 (2022)
Publisher : LPPM Universitas Bumigora

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/adma.v2i2.1568

Abstract

Thinking about computing or Computational Thinking (CT) is one of the problem solving techniques that is very important today to prepare the next generation to be competitive in this digital economy era. This skill teaches students how to think the way scientists think, to solve real-world problems. In training or familiarizing children with CT, this can be done by including or adding CT to the learning strategy. The intended strategy can be related to the packaging of the material, the learning media used, or an interesting learning model that can familiarize children with CT. The learning media used are expected to help children understand the subject matter. For early childhood, learning media has a vital role in learning. This is because at that age children are still at the stage of learning while playing. On August 6, 2019, Bumigora University signed a charter of cooperation with the coordination of the national committee (National Board Organization (NBO) bebras Indonesia), where for the next 5 (five) years Bumigora University will become a partner in realizing PANDAI students . PANDAI, is the name chosen as the movement to socialize CT, which stands for Pengjar Digital Era Indonesia. The PANDAI Movement training in collaboration with Bumigora University is also the first teacher training to be held in the entire West Nusa Tenggara region, involving approximately 400 elementary and junior high school teachers. Service activities related to Computational Thinking have been carried out. The results of the implementation have been evaluated with the results that most of the teacher participants agreed to apply the results of the training related to Computational Thinking to students and also most of the participants agreed to join the follow-up programs from Bebras. The implementation stage for students is still unable to carry out the service team due to the current pandemic situation, but with the continuation of the activities carried out by the UBG bureau with NBO Bebras. It is hoped that this activity can run continuously and be supported positively by the parties involved.