Erick Purwanto

Dr. Erick Purwanto

Department of Computer Science, National University of Singapore, Singapore City, Singapore
orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 July 2024 in Sensors
Verified
Journal Article
Open access
PDF file
Haoyu WuHaoyu Wu
Haoyu Wu
JQJun Qi
Jun Qi
EPErick Purwanto
Erick Purwanto
XZXiaohui Zhu
Xiaohui Zhu
PYPo Yang
Po Yang
JCJianjun Chen
Jianjun Chen
Objective: Motivated by Health Care 4.0, this study aims to reducing the dimensionality of traditional EEG features based on manual extracted features, including statistical features in the time and frequency domains. Methods: A total of 22 multi-scale features were extracted from the UNM and Iowa datasets using a 4th order Butterworth filter and wavelet packet transform. Based on single-channel validation, 29 channels with the highest R2 scores were selected from a pool of 59 common channels. The proposed channel selection scheme was validated on the UNM dataset and tested on the Iowa dataset to compare its generalizability against models trained without channel selection. Results: The experimental results demonstrate that the proposed model achieves an optimal classification accuracy of 100%. Additionally, the generalization capability of the channel selection method is validated through out-of-sample testing based on the Iowa dataset Conclusions: Using single-channel validation, we proposed a channel selection scheme based on traditional statistical features, resulting in a selection of 29 channels. This scheme significantly reduced the dimensionality of EEG feature vectors related to Parkinson’s disease by 50%. Remarkably, this approach demonstrated considerable classification performance on both the UNM and Iowa datasets. For the closed-eye state, the highest classification accuracy achieved was 100%, while for the open-eye state, the highest accuracy reached 93.75%.

26 May 2025 in IEEE Transactions on Education
Verified
Journal Article
Closed access
TMTeng Ma
Teng Ma
Ahmad Samed Al-AdwanAhmad Samed Al-Adwan
Ahmad Samed Al-Adwan
Na LiNa Li
Na Li
EPErick Purwanto
Erick Purwanto
WMWan Meng
Wan Meng
LHLiang Hai-Ning
Liang Hai-Ning
Contribution: This study has proposed a hybrid framework of acceptance and self-determination for the use of digital textbooks in higher education programming courses. The intertwined relationships between acceptance and self-determination factors, and their joint effects on student’s engagement and learning performance are all examined. Background: Despite the proliferation of education technologies, their acceptance and performance among students are still limited by the active participation of students, especially for engineering training of higher education. Traditional acceptance-related factors face challenges in explaining the engagement of students in using new technology, which calls for an integrated examination for their interconnection and joint effects. Research Questions: 1) How do acceptance and self-determination factors influence each other in using digital textbooks? 2) How do acceptance and self-determination factors jointly affect student’s engagement and learning performance? Methodology: By collecting data from students in a programming course using HTML5 package (H5P) digital textbooks, this study examined relationships in the hybrid framework through partial least-squares structural equation modeling (PLS-SEM). Findings: This study reveals the relationship between acceptance and self-determination, which verified the mediation effect of self-determination between perceived usefulness / ease of use and acceptance. Besides, the acceptance of students, along with competence and relatedness, could significantly facilitate learning engagement, which further increases the learning performance.

20 January 2025 in IEEE Transactions on Intelligent Transportation Systems
Verified
Journal Article
Closed access
Runwei GuanRunwei Guan
Runwei Guan
LJLiye Jia
Liye Jia
SYShanliang Yao
Shanliang Yao
FYFengyufan Yang
Fengyufan Yang
SXSheng Xu
Sheng Xu
YYYutao Yue
Yutao Yue
Waterway perception is critical for the special operations and autonomous navigation of Unmanned Surface Vessels (USVs), but current perception schemes are sensor-based, neglecting the interaction between humans and USVs for embodied perception in various operations. Therefore, inspired by visual grounding, we present WaterVG, the inaugural visual grounding dataset tailored for USV-based waterway perception guided by human prompts. WaterVG contains a wealth of prompts describing multiple targets, with instance-level annotations, including bounding boxes and masks. Specifically, WaterVG comprises 11,568 samples and 34,987 referred targets, integrating both visual and radar characteristics. The text-guided two-sensor pattern provides a fine granularity of text prompts aligned with the visual and radar features of the referent targets, containing both qualitative and numeric descriptions. To enhance the endurance and maintain the normal operations of USVs in open waterways, we propose Potamoi, a low-power visual grounding model. Potamoi is a multi-task model employing a sophisticated Phased Heterogeneous Modality Fusion (PHMF) mechanism, which includes Adaptive Radar Weighting (ARW) and Multi-Head Slim Cross Attention (MHSCA). The ARW module utilizes a gating mechanism to adaptively extract essential radar features for fusion with visual inputs, ensuring prompt alignment. MHSCA, characterized by its low parameter count and computational efficiency (FLOPs), effectively integrates contextual information from both sensors with linguistic features, delivering outstanding performance in visual grounding tasks. Comprehensive experiments and evaluations on WaterVG demonstrate that Potamoi achieves state-of-the-art results compared to existing methods. The project is available at https://github.com/GuanRunwei/WaterVG.

10 October 2022 in Interactive Learning Environments
Verified
Journal Article
Closed access
Na LiNa Li
Na Li
EPErick Purwanto
Erick Purwanto
XZXiaojun Zhang
Xiaojun Zhang
FCFeng Cao
Feng Cao
KWKok Hoe Wong
Kok Hoe Wong
XCXiangru Chen
Xiangru Chen
To address the problem of poor efficacy in hybrid learning in higher education, this study conducted a mix-method explanatory case study with structural equation modelling to examine a new construct, “perceived pedagogical value” (PPV), and its relationship with the core constructs of the Unified Theory of Acceptance and Use of Technology (UTAUT) model. This study recruited 246 undergraduate students from a year 2 computer science course at an English Medium Instruction University in China to investigate their perceptions of using JazzQuiz for interactive hybrid learning using an online survey and semi-structured interviews. One of the key findings suggests that students who have a deeper understanding of technology’s pedagogical value (i.e. higher PPV) are more likely to accept and use the technology with reduced efforts through five dimensions. Our contribution to the literature is threefold: (1) the new construct, PPV, extends our understanding of the UTAUT model for interactive hybrid learning; (2) the extended model highlights the role of student PPV in improving learning efficacy; (3) the five dimensions of how PPV influences technology acceptance, and student academic performance clarify the underlying mechanisms of effective interactive hybrid learning. Further studies are needed to test the new model in different contexts.

11 February 2025 in SoftwareX
Verified
Journal Article
Open access
MLMingyuan Li
Mingyuan Li
DWDuan Wang
Duan Wang
EPErick Purwanto
Erick Purwanto
TSThomas Selig
Thomas Selig
QZQing Zhang
Qing Zhang
HLHai-Ning Liang
Hai-Ning Liang

24 September 2024
Verified
Conference Paper
Open access
TWTianyi Wu
Tianyi Wu
YHYongrun Huang
Yongrun Huang
EPErick Purwanto
Erick Purwanto
PCPaul Craig
Paul Craig
This study presents an Emotion Recognition Multi-Attention Model (EmoMA-Net), a novel multimodal neural network aimed at enhancing real-time emotion recognition in educational environments. By leveraging the WESAD dataset, our model combines Convolutional Neural Networks (CNN), a Time Series Memory System (TSMS), and a Multi-Attention Mechanism to analyze diverse physiological signals, such as heart rate variability (HRV) and electroencephalogram (EEG). Unlike traditional emotion recognition methods reliant on subjective self-reports, our model delivers objective and accurate predictions of student stress levels through multimodal physiological data collected from wearable sensors. Achieving accuracy up to 99.66%, it facilitates adaptive educational systems to provide real-time feedback to educators, enabling prompt adjustments to teaching strategies. This advancement represents forward in emotion prediction technology, contributing to more responsive and adaptive educational experiences based on real-time emotional insights.