Article Per Year (5 Year)
p-Index From 2021 - 2026
2.457
P-Index
This Author published in this journals
All Journal Jurnal Infra
Liliana Liliana
Program Studi Teknik Informatika, Universitas Kristen Petra Surabaya
Computer Science & IT

Published : 19 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 19 Documents
Search

 1   2 
Pewarnaan Otomatis Sketsa Gambar Menggunakan Metode Conditional GAN Untuk Mempercepat Proses Pewarnaan Regan Reinaldo Kalendesang; Liliana Liliana; Djoni Haryadi Setiabudi
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

Anime is a Japanese animation that consists of many frames of images. Images that used to make an anime can be made using hand-drawn or using digital-drawn. It takes a lot of time to make an anime. In making anime for 1 second, it needs a total of 24 frames, this is why it takes a lot of time to make anime and also takes a lot of money. Each image also needs to be colored, this is also why making anime takes so much time. The method used in this research is GAN (Generative Adversarial Network) or should we call C-GAN (Conditional Generative Adversarial Network) to make coloring anime sketches easier. Dataset that is used in this research is a pair of sketch images and sketch images that have already been colored.
Penerapan 3D Human Pose Estimation Indoor Area untuk Motion Capture dengan Menggunakan YOLOv4-Tiny, EfficientNet Simple Baseline, dan VideoPose3D Gerry Steven; Liliana Liliana; Anita Nathania Purbowo
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

Human pose estimation is a research topic that has goal to estimate every human’s keypoint coordinate that can be connected and make a human skeleton. The development of this topic can be applicated to human activity recognition, human tracking, and motion capture for film and animation. There are several challenges for this topic: diverse human pose, diverse body appearance from clothing and similar parts, and complex environment that may cause foreground occlusion. There are several methods to be used in this research: YOLOv4- Tiny, EfficientNet Simple Baseline, and VideoPose3D. YOLOv4- Tiny will process image input to get bounding box coordinate. This coordinate will be inputted to EfficientNet Simple Baseline modification to get 16 keypoint 2D coordinates. After that, VideoPose3D will processed 2D coordinates into 15 keypoints 3D coordinates. The result from this research is EfficientNet Simple Baseline modification is faster with 4.54ms time compared to its original with time of 5.15ms. Although faster, its modification has its own downside. In term of accuracy, modification still less accurate than its original with highest average Percentage of Correct Keypoints head (PCKh@0.2) 86.89%, and original with PCKh@0.2 89.62%. This affect 3D human pose estimation using VideoPose3D, where using EfficientNet modification resulting Mean Per Joints Position Error (MPJPE) 25.3 mm compared to original Simple Baseline resulting MPJPE 28.1mm.
Penerapan metode hand gesture recognition dalam melakukan kontrol terhadap aplikasi powerpoint dan media player untuk kebutuhan online conference William Sean Wiyogo; Liliana Liliana
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

Since the prolonged COVID-19 pandemic, most human activities are seen with the concept of virtual meetings. This concept is helped by the use of an online conferencing platform. As a result, needs arise among the new society. The learning model of hybrid learning, or blended learning is a combination of face-to-face learning with e-learning. This learning method reduces teaching performance due to limited range of motion. Thus, hand tracking gesture recognition can be used as a solution to overcome this problem. This study aims to model a gesture recognition system with statistical and dynamic recognition. The method used in this research is CNN-based RT3D_16F which is used as dynamic motion prediction and Mediapipe hand pipeline which is used as static motion prediction. The data set used consists of 27 movement labels (includes 2 movement labels that shouldn't be recognized as specific moves).
Deteksi Plagiarisme pada Kode Bahasa Pemrograman Java menggunakan XGBoost Tomy Widjaja; Andre Gunawan; Liliana Liliana
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

With the ease of access to information and cloud server technology, it makes it easier for anyone to access the code data. Coupled with the industry 4.0 era, the number of informatics students is also increasing rapidly. This makes code plagiarism easier to do, especially in academic environment Manual checking of plagiarism is repetitive, difficult, and time-consuming task. Therefore, automation for high quality source code plagiarism detection is needed. The dataset used in this research was collected from “Dasar Pemrograman” class at Petra Christian University. After that the code will continue to tokenization preprocessing using java grammar stage. Then, the algorithm will calculate pairwise features using 3 main algorithms, namely levenshtein distance, greedy string tiling, and bigram which will produce 12 features and a collection of statistic features. Finally, the features will be used for the training and inference process on the XGBoost model. The test result shows that the proposed features have better performance metrics than previous research, it has f1-score of 99%. Implementation of preprocessing can also improve performance metrics on the features proposed in this study and in previous research.
Pengenalan Rambu Lalu Lintas di Indonesia Secara Realtime Menggunakan YOLOv4-tiny Gregorius Nicholas Goenawan; Alvin Nathaniel Tjondrowiguno; Liliana Liliana
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

Concentration are crucial when driving. Drivers who lose their concentration tend to have a slower reaction time, and a higher possibility of violating traffic signs. Traffic signs violation is considered a criminal act with harsh penalties. In addition, traffic sign violations interferes with comfort and endanger other road users. Therefore, we need a system that is able to detect signs accurately and quickly which can inform driver in advance. A research on traffic signs detection on Swedish and Slovenian traffic signs use Mask R-CNN model which based on convolutional neural networks [18]. These method was capable of achieving a mAP@50 score that exceeds 95%. However, the research did not evaluate on the detection speed of such methods. In this research, YOLOv4-tiny is used to detect Indonesian traffic signs. Dataset used in this research are independently collected, which consist of nine prohibition signs and two command signs. The YOLOv4-tiny method with input size of 416 x 416 is able to achieve mAP@50 score of 88.55% with detection speed of 19.41 FPS. With modification to input size and dataset, YOLOv4-tiny are able to achieve mAP@50 score up to 89.58% and detection speed up to 30.87 FPS. YOLOv4-tiny are also able to detect road signs from distance of around 5 to 15 meters with 80.42 % accuracy. Indonesian traffic sign recognition program made by utilizing the YOLOv4-tiny model achieve average recall of 72.9%.
Penerapan SVM untuk Klasifikasi Sentimen pada Review Comment Berbahasa Indonesia di Online Shop Yoshua Refo; Silvia Rostianingsih; Liliana Liliana
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

With so many users accessing online shops, comments are an important aspect when shopping. Buyers can provide comments about the goods or thing that have been purchased, both negative comments and positive comments. By collecting various kinds of comments, the data can be used to classify comments. This research will use the Support Vector Machine (SVM) algorithm which is considered the right method for text classification. The method will be tested for its performance, seen from how good and accurate the method used in classifying comments is. In addition, this research also uses kernels, namely Linear kernels, Radial Basis Function (RBF) kernels, and Polynomial kernels as test scenarios. Based on the test results shown, SVM is a good method in classifying text. SVM classifies text that has gone through the preprocessing stage with an accuracy value of 88% on the RBF kernel, 87% on the linear kernel, and 87% on the polynomial kernel. The accuracy value in the aspect classification itself is 78% on the RBF kernel, 78% on the Linear kernel and 74% on the Polynomial kernel.
Form Evaluasi Online Mata Kuliah Pra Skripsi Dan Skripsi Berbasis Android Stephen Cornelius Hertanto; Silvia Rostianingsih; Liliana Liliana
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

In order to support the vision of "Green Campus" implemented by Petra Christian University Surabaya as well as to accommodate the evaluation interests of courses in the new curriculum, namely pre-thesis and thesis. At the evaluation stage of the course, it involves many parties, so to digitize all documents and facilitate coordination, an application is needed to assist the entry process and produce the required reports. So far, the evaluation process that has occurred is quite long and takes a lot of time, namely when evaluating the Pre-Thesis to Thesis. As well as seeing the increasing number of Petra students every year, the coordinating lecturers and supervisors on Pre Thesis and Thesis will be more and more difficult in handling the evaluation. Therefore, this thesis creates an "Online Evaluation Form for Pre-thesis and Thesis courses based on Android" which is an application on an Android-based smartphone that aims to facilitate the coordination of lecturers and supervisors in evaluating these courses. By making this application, it will be easier for the lecturers to fill in the scores and evaluation of the Pre-thesis and Thesis reports without taking much time, and the data created can be more accurate and faster. So that the delivery of grades to students becomes shorter.
Aplikasi Penerjemah Kegiatan Seminar Menjadi Video Bahasa Isyarat BISINDO Dengan Speech To Text Marcel Slamet Sugianto; Liliana Liliana; Anita Nathania Purbowo
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

Information at this time is very much needed to increase our knowledge. However, this delivery can be hindered by several conditions such as the inability to hear the deaf community. Based on data from the Data and Information Center of the Ministry of Health of the Republic of Indonesia in 2019, 7.09% of the Indonesian population is deaf. In addition, the delivery of information at the seminar can be hindered by noise and participants sitting far from the speaker will have difficulty hearing the speaker's voice. In this study, we will use Speech-To-Text on the Android application which aims to help translate information in the form of voice delivered as at a seminar into text and will be converted into BISINDO sign language video. The results of testing the use of the Speech-To-Text feature in the application that has been made show that it is able to accommodate approximately 100 words in 1 minute at a time when the speaker speaks without any pause. The Speech-To-Text feature used takes approximately 2 seconds to translate the received voice and the time lag required by the speaker device to the participant's device takes approximately 3-5 seconds after using 5 different internet speeds. For the accuracy of the Speech-To-Text feature that was tested using 3 narrations read by 4 different people, the accuracy of the Speech-To-Text feature has an accuracy of above 80% in general, although there is an accuracy that is below 80% due to the ambiguity of the pronunciation.
Sistem Backend dari Aplikasi Mobile dan Website untuk Sistem Registrasi, Reservasi, dan Identifikasi Penumpang Shuttle Bus UK Petra Michelle Christiana Chandra; Rolly Intan; Liliana Liliana
Jurnal Infra Vol 10, No 2 (2022)
Publisher : Universitas Kristen Petra

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

Abstract

In 2022, Petra Christian University (PCU) launched its shuttle bus program exclusive to the academic community of PCU. This bus operates by picking up and dropping off passengers on the route of West Surabaya – PCU and vice versa. There are overwhelmingly many more members of the academic community of PCU residing in West Surabaya compared to the capacity of the shuttle bus. Registration, reservation, and identification systems are necessary to resolve this problem. The registration system ensures that the passengers are members of the academic community of PCU. The reservation system supports passengers in booking seats for the desired trips. The identification system authenticates the passengers when boarding the shuttle bus. The PCU shuttle bus program is operational and has been using the deployed application properly. Every function and API in the application passes the designed feature testing and matches the appropriate expectations. A survey was conducted of 76 passengers and 1 admin and 2 drivers were interviewed. From 76 passenger respondents, the average satisfaction score is 4.259 out of 5. From the 2 driver interviews, the application received a score of 8 and 10 out of 10. And lastly, the average satisfaction for the admin system and website is 4.75 out of 5.
Co-Authors Agustinus Noertjahyana Alvin Nathaniel Tjondrowiguno Andre Gunawan Andrew Firman Saputra Anita Nathania Purbowo Djoni Haryadi Setiabudi Franky Halim Gerry Steven Gregorius Nicholas Goenawan Hans Juwiantho Henry Novianus Palit Henry Wicaksono Ivana Jovita Handoko Kartika Gunadi Kevin Jonathan Kevin Reynaldi Tanjung Lois Fernando Audi Marcel Slamet Sugianto Michael Alexander Rustan Michael Budiono Michelle Christiana Chandra Regan Reinaldo Kalendesang Rolly Intan Silvia Rostianingsih Stephen Cornelius Hertanto Tomy Widjaja William Sean Wiyogo Willy Pratama Darmalim Yoshua Refo