DLDongjin Lee

Dongjin Lee

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Dongjin Lee has a total of 23 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Cross-Domain Evaluation and Fine-Tuned Adaptation of iCatcher+ for Korean Infant Gaze Data

    by Rajalakshmi Madhavan, Jiho Lee, Dongjin Lee, Jae-Hun Jung, Eon-Suk Ko

    This study evaluates the cross-domain generalisability of three iCatcher+ gaze classification models on a dataset of Korean infants. Applying the pretrained models to the Korean dataset resulted in a substantial decline in accuracy across all models. Performance was systematically modulated by developmental mismatch: accuracy declined as the age distribution of the target dataset diverged from that of the training data. In contrast, experiment-level factors such as stimulus type and temporal progression had limited effects. Fine-tuning the Lookit model on Korean data improved target-domain performance, but at the cost of reduced accuracy on the original source dataset, consistent with catastrophic forgetting. Our findings highlight that deep learning-based gaze classifiers remain sensitive to domain shift. Although fine-tuning can partially mitigate performance loss, robust generalisation requires closer alignment between training and deployment contexts.

  • Open Access

    Investigating the Effects of Point Source Injection Strategies on KMTNet Real/Bogus Classification

    by Dongjin Lee, Gregory S. H. Paek, Seo-Won Chang, Changwan Kim, Mankeun Jeong, Hongjae Moon, Seong-Heon Lee, Jae-Hun Jung, Myungshin Im

    Recently, machine learning-based real / bogus (RB) classifiers have demonstrated effectiveness in filtering out artifacts and identifying genuine transients in real-time astronomical surveys. However, the rarity of transient events and the extensive human labeling required for a large number of samples pose significant challenges in constructing training data sets for RB classification. Given these challenges, point source injection techniques, which inject simulated point sources into optical images, provide a promising solution. This paper presents the first detailed comparison of different point source injection strategies and their effects on classification performance within a simulation-to-reality framework. To this end, we first construct various training data sets based on random injection (RI), near galaxy injection (NGI), and a combined approach by using the Korea Microlensing Telescope Network data sets. Subsequently, we train convolutional neural networks on simulated cutout samples and evaluate them on real, imbalanced data sets from gravitational wave follow-up observations for GW190814 and S230518h. Extensive experimental results show that RI excels at asteroid detection and bogus filtering but underperforms on transients occurring near galaxies (e.g., supernovae). In contrast, NGI is effective for detecting transients near galaxies but tends to misclassify variable stars as transients, resulting in a high false positive rate. The combined approach effectively handles these trade-offs, thereby balancing between detection rate and false positive rate. Our results emphasize the importance of point source injection strategy in developing robust RB classifiers for transient (or multimessenger) follow-up campaigns.

  • Open Access

    Investigating the Effects of Point Source Injection Strategies on KMTNet Real/Bogus Classification

    by Dongjin Lee, Gregory S. H. Paek, Seo-Won Chang, Changwan Kim, Mankeun Jeong, Hongjae Moon, Seong-Heon Lee, Jae-Hun Jung, Myungshin Im

    Recently, machine learning-based real/bogus (RB) classifiers have demonstrated effectiveness in filtering out artifacts and identifying genuine transients in real-time astronomical surveys. However, the rarity of transient events and the extensive human labeling required for a large number of samples pose significant challenges in constructing training datasets for RB classification. Given these challenges, point source injection techniques, which inject simulated point sources into optical images, provide a promising solution. This paper presents the first detailed comparison of different point source injection strategies and their effects on classification performance within a simulation-to-reality framework. To this end, we first construct various training datasets based on Random Injection (RI), Near Galaxy Injection (NGI), and a combined approach by using the Korea Microlensing Telescope Network datasets. Subsequently, we train convolutional neural networks on simulated cutout samples and evaluate them on real, imbalanced datasets from gravitational wave follow-up observations for GW190814 and S230518h. Extensive experimental results show that RI excels at asteroid detection and bogus filtering but underperforms on transients occurring near galaxies (e.g., supernovae). In contrast, NGI is effective for detecting transients near galaxies but tends to misclassify variable stars as transients, resulting in a high false positive rate. The combined approach effectively handles these trade-offs, thereby balancing between detection rate and false positive rate. Our results emphasize the importance of point source injection strategy in developing robust RB classifiers for transient (or multi-messenger) follow-up campaigns.

  • Open Access

    GECKO Follow-up Observations of the Binary Neutron Star–Black Hole Merger Candidate S230518h

    by Gregory S. H. Paek, Myungshin Im, Mankeun Jeong, Seo-Won Chang, Martin Moonkuk Hur, YoungPyo Hong, Sophia Kim, Jaewon Lee, Dongjin Lee, Seong-Heon Lee, et al.

    The gravitational-wave (GW) event S230518h is a potential binary neutron star–black hole merger (NSBH) event that was detected during engineering run 15, which served as the commissioning period before the LIGO–Virgo–KAGRA O4a observing run. Despite its low probability of producing detectable electromagnetic emissions, we performed extensive follow-up observations of this event using the Gravitational-wave Electromagnetic Counterpart Korean Observatories (GECKO) telescopes in the Southern Hemisphere. Our observations covered 61.7% of the 90% credible region, a 284 deg2 area accessible from the Southern Hemisphere, reaching a median limiting magnitude of R = 21.6 mag. In these images, we conducted a systematic search for an optical counterpart of this event by combining a convolutional-neural-network-based classifier and human verification. We identified 128 transient candidates, but no significant optical counterpart was found that could have caused the GW signal. Furthermore, we provide feasible kilonova properties that are consistent with the upper limits of the observations. Although no optical counterpart was found, our result demonstrates both GECKO’s efficient wide-field follow-up capabilities and usefulness for constraining properties of kilonovae from NSBH mergers at distances of ∼200 Mpc.


  • Open Access

    The effects of topological features on convolutional neural networks—an explanatory analysis via Grad-CAM

    by Dongjin Lee, Seong-Heon Lee, Jae-Hun Jung

    Topological data analysis (TDA) characterizes the global structure of data based on topological invariants such as persistent homology, whereas convolutional neural networks (CNNs) are capable of characterizing local features in the global structure of the data. In contrast, a combined model of TDA and CNN, a family of multimodal networks, simultaneously takes the image and the corresponding topological features as the input to the network for classification, thereby significantly improving the performance of a single CNN. This innovative approach has been recently successful in various applications. However, there is a lack of explanation regarding how and why topological signatures, when combined with a CNN, improve discriminative power. In this paper, we use persistent homology to compute topological features and subsequently demonstrate both qualitatively and quantitatively the effects of topological signatures on a CNN model, for which the Grad-CAM analysis of multimodal networks and topological inverse image map are proposed and appropriately utilized. For experimental validation, we utilize two famous datasets: the transient versus bogus image dataset and the HAM10000 dataset. Using Grad-CAM analysis of multimodal networks, we demonstrate that topological features enforce the image network of a CNN to focus more on significant and meaningful regions across images rather than task-irrelevant artifacts such as background noise and texture.