Byungseok Min
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Byungseok Min has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Tae-Hyuk Ahn
Department of Computer Science, Saint Louis University, St. Louis, MO 63103, USA
4 shared publications - Gihwan Jung
Department of Computer Science, Saint Louis University, Saint Louis, MO 63103, USA
2 shared publications Aendri Dhiren
Program of Artificial Intelligence, Saint Louis University, St. Louis, MO, USA
1 shared publication- Cory Gardner
Department of Computer Science Saint Louis University 63103 St. Louis MO USA
1 shared publication - Cory Gardner
Department of Computer Science, Saint Louis University, St. Louis, MO, USA
1 shared publication - Jazzson Park
SAVE EARTH, SAVE US, Sesu.ai, Seongnam, South Korea
1 shared publication
Latest Publications
- Open Access
- 31 May 2026
- Published by The Korean Institute of Broadcast and Media Engineers in Journal of Broadcast Engineering
- Vol. 31 (3) , 373-392
An Enhanced Keyframe Extraction Technique based on Multi-Object Tracking and Re-Identification
- 31 May 2026
- Published by The Korean Institute of Broadcast and Media Engineers in Journal of Broadcast Engineering
- Vol. 31 (3) , 373-392
Open AccessAn Enhanced Keyframe Extraction Technique based on Multi-Object Tracking and Re-Identification
- Open Access
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 37702-37718
Wavelet-Enhanced PaDiM for Industrial Anomaly Detection
Anomaly detection (AD) and localization in industrial images are essential for automated quality inspection. PaDiM models normality using multivariate Gaussian distributions over deep features, but typically relies on random channel selection for dimensionality reduction. This limits interpretability and potentially discards structured information. We propose Wavelet-Enhanced PaDiM (WE-PaDiM), which replaces random selection with a structured frequency-domain feature selection strategy using the 2D Discrete Wavelet Transform (DWT) as its core component to decompose CNN features into interpretable frequency subbands before modeling. Specifically, DWT is applied independently to multi-layer CNN feature maps, selected subband coefficients are spatially aligned, and the resulting representations are concatenated prior to PaDiM modeling. We evaluate WE-PaDiM on MVTec AD and VisA datasets across ResNet-18 and EfficientNet-b0–b6 backbones. Under per-category configuration selection, WE-PaDiM achieves 99.32% Image-AUC on MVTec AD (image-optimized) and 96.65% Pixel-AUC (pixel-optimized), and achieves 93.59% Image-AUC on VisA (image-optimized) and 96.43% Pixel-AUC (pixel-optimized). These image- and pixel-optimized results correspond to distinct operating points (potentially selecting different configurations), rather than a single jointly-achieved setting. Notably, the dominant fixed pixel-level operating point is Haar/L1/LL (low-frequency) on both datasets, while per-category pixel-optimal selections are largely LL-dominant with a small number of VisA exceptions. WE-PaDiM offers deterministic and interpretable feature selection with competitive computational characteristics.
- 3 March 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 14, 37702-37718
Open AccessWavelet-Enhanced PaDiM for Industrial Anomaly Detection
Anomaly detection (AD) and localization in industrial images are essential for automated quality inspection. PaDiM models normality using multivariate Gaussian distributions over deep features, but typically relies on random channel selection for dimensionality reduction. This limits interpretability and potentially discards structured information. We propose Wavelet-Enhanced PaDiM (WE-PaDiM), which replaces random selection with a structured frequency-domain feature selection strategy using the 2D Discrete Wavelet Transform (DWT) as its core component to decompose CNN features into interpretable frequency subbands before modeling. Specifically, DWT is applied independently to multi-layer CNN feature maps, selected subband coefficients are spatially aligned, and the resulting representations are concatenated prior to PaDiM modeling. We evaluate WE-PaDiM on MVTec AD and VisA datasets across ResNet-18 and EfficientNet-b0–b6 backbones. Under per-category configuration selection, WE-PaDiM achieves 99.32% Image-AUC on MVTec AD (image-optimized) and 96.65% Pixel-AUC (pixel-optimized), and achieves 93.59% Image-AUC on VisA (image-optimized) and 96.43% Pixel-AUC (pixel-optimized). These image- and pixel-optimized results correspond to distinct operating points (potentially selecting different configurations), rather than a single jointly-achieved setting. Notably, the dominant fixed pixel-level operating point is Haar/L1/LL (low-frequency) on both datasets, while per-category pixel-optimal selections are largely LL-dominant with a small number of VisA exceptions. WE-PaDiM offers deterministic and interpretable feature selection with competitive computational characteristics.
- Open Access
Early Wildfire Smoke Detection with a Multi-Resolution Framework and Two-Stage Classification Pipeline
Early wildfire smoke detection is critical for preventing small ignitions from escalating into large-scale fires, yet early-stage smoke plumes are often faint, low-contrast, and spatially small. When full-resolution frames are resized to satisfy fixed-input detector architectures and enable efficient batched GPU inference, these subtle cues are further diminished, leading to missed detections and unreliable scores near deployment thresholds. Existing remedies such as multi-scale inference, slicing/tiling, or super-resolution could improve sensitivity, but typically incur substantial overhead from multiple forward passes or added network components, limiting real-time use on resource-constrained platforms. To mitigate these challenges, we propose a composite multi-resolution detection framework that improves sensitivity to small smoke regions while maintaining single-pass inference. Motivated by the fact that most operational wildfire monitoring systems rely on Unmanned Aerial Vehicle (UAV) platforms and mountain-top Closed-Circuit Television (CCTV) systems surveillance, their wide-field imagery typically contains a large sky region above the horizon where early smoke is most likely to first become visible. Accordingly, crop placement is guided by a skyline prior that prioritizes this high-probability sky band while retaining the remaining scene for global context. A dynamic compositing stage stacks a global view with a high-resolution, sky-aligned band into a standard square detector input, preserving context with minimal added cost. Detections from the two views are reconciled via coordinate restoration and non-maximum suppression. For deployment, a lightweight second-stage classifier selectively re-evaluates low-confidence detections to stabilize decisions near a fixed operating threshold without retraining the detector. Compared to the baseline detector, our approach improves detection performance on the Early Smoke dataset, achieving gains of +4.6 percentage points in AP @0.5:0.95, +3.4 percentage points in AP @0.5, +2.9 percentage points in precision, +5.3 percentage points in recall, and +4.3 percentage points in F1-score.
Open AccessEarly Wildfire Smoke Detection with a Multi-Resolution Framework and Two-Stage Classification Pipeline
Early wildfire smoke detection is critical for preventing small ignitions from escalating into large-scale fires, yet early-stage smoke plumes are often faint, low-contrast, and spatially small. When full-resolution frames are resized to satisfy fixed-input detector architectures and enable efficient batched GPU inference, these subtle cues are further diminished, leading to missed detections and unreliable scores near deployment thresholds. Existing remedies such as multi-scale inference, slicing/tiling, or super-resolution could improve sensitivity, but typically incur substantial overhead from multiple forward passes or added network components, limiting real-time use on resource-constrained platforms. To mitigate these challenges, we propose a composite multi-resolution detection framework that improves sensitivity to small smoke regions while maintaining single-pass inference. Motivated by the fact that most operational wildfire monitoring systems rely on Unmanned Aerial Vehicle (UAV) platforms and mountain-top Closed-Circuit Television (CCTV) systems surveillance, their wide-field imagery typically contains a large sky region above the horizon where early smoke is most likely to first become visible. Accordingly, crop placement is guided by a skyline prior that prioritizes this high-probability sky band while retaining the remaining scene for global context. A dynamic compositing stage stacks a global view with a high-resolution, sky-aligned band into a standard square detector input, preserving context with minimal added cost. Detections from the two views are reconciled via coordinate restoration and non-maximum suppression. For deployment, a lightweight second-stage classifier selectively re-evaluates low-confidence detections to stabilize decisions near a fixed operating threshold without retraining the detector. Compared to the baseline detector, our approach improves detection performance on the Early Smoke dataset, achieving gains of +4.6 percentage points in AP @0.5:0.95, +3.4 percentage points in AP @0.5, +2.9 percentage points in precision, +5.3 percentage points in recall, and +4.3 percentage points in F1-score.
- Open Access
- 13 January 2026
- Published by MDPI AG in Preprints.org
Early Wildfire Smoke Detection with a Multi-Resolution Framework and Two-Stage Classification Pipeline
Early wildfire smoke detection is critical for preventing small ignitions from escalating into large-scale fires, yet early-stage smoke plumes are often faint, low-contrast, and spatially small. When full-resolution frames are resized to satisfy fixed-input detector architectures and enable efficient batched GPU inference, these subtle cues are further diminished, leading to missed detections and unreliable scores near deployment thresholds. Existing remedies such as multi-scale inference, slicing/tiling, or super-resolution could improve sensitivity, but typically incur substantial overhead from multiple forward passes or added network components, limiting real-time use on resource-constrained platforms. To mitigate these challenges, we propose a composite multi-resolution detection framework that improves sensitivity to small smoke regions while maintaining single-pass inference. Motivated by the fact that most operational wildfire monitoring systems rely on UAV platforms and mountain-top CCTV surveillance, their wide-field imagery typically contains a large sky region above the horizon where early smoke is most likely to first become visible. Accordingly, crop placement is guided by a skyline prior that prioritizes this high-probability sky band while retaining the remaining scene for global context. A dynamic compositing stage stacks a global view with a high-resolution, sky-aligned band into a standard square detector input, preserving context with minimal added cost. Detections from the two views are reconciled via coordinate restoration and non-maximum suppression. For deployment, a lightweight second-stage classifier selectively re-evaluates low-confidence detections to stabilize decisions near a fixed operating threshold without retraining the detector. Across Early Smoke and Pyro-SDIS, the method improves AP@0.5 by +0.060/+0.036 and APsmall by +0.151/+0.025, respectively, with accuracy-efficiency trade-offs supported by ablation studies and deployment-time latency analysis.
- 13 January 2026
- Published by MDPI AG in Preprints.org
Open AccessEarly Wildfire Smoke Detection with a Multi-Resolution Framework and Two-Stage Classification Pipeline
Early wildfire smoke detection is critical for preventing small ignitions from escalating into large-scale fires, yet early-stage smoke plumes are often faint, low-contrast, and spatially small. When full-resolution frames are resized to satisfy fixed-input detector architectures and enable efficient batched GPU inference, these subtle cues are further diminished, leading to missed detections and unreliable scores near deployment thresholds. Existing remedies such as multi-scale inference, slicing/tiling, or super-resolution could improve sensitivity, but typically incur substantial overhead from multiple forward passes or added network components, limiting real-time use on resource-constrained platforms. To mitigate these challenges, we propose a composite multi-resolution detection framework that improves sensitivity to small smoke regions while maintaining single-pass inference. Motivated by the fact that most operational wildfire monitoring systems rely on UAV platforms and mountain-top CCTV surveillance, their wide-field imagery typically contains a large sky region above the horizon where early smoke is most likely to first become visible. Accordingly, crop placement is guided by a skyline prior that prioritizes this high-probability sky band while retaining the remaining scene for global context. A dynamic compositing stage stacks a global view with a high-resolution, sky-aligned band into a standard square detector input, preserving context with minimal added cost. Detections from the two views are reconciled via coordinate restoration and non-maximum suppression. For deployment, a lightweight second-stage classifier selectively re-evaluates low-confidence detections to stabilize decisions near a fixed operating threshold without retraining the detector. Across Early Smoke and Pyro-SDIS, the method improves AP@0.5 by +0.060/+0.036 and APsmall by +0.151/+0.025, respectively, with accuracy-efficiency trade-offs supported by ablation studies and deployment-time latency analysis.
- Open Access
- 8 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. PP, 1
LSTM-Based Online Learning for Real-Time Solar Power Forecasting
Accurate solar power forecasting is essential to mitigate the variability and uncertainty of solar power generation and integrate renewable energy sources into power grids. A multi-step Long Short Term Memory (LSTM)-based model is proposed for solar power generation forecasting using online learning. To capture daily fluctuations in solar energy patterns, the model employs a dynamic sunrise–sunset data window operation, ensuring that the relevant transitions in solar irradiance are accurately represented in the training set. For seasonality analysis, the variance of the correlation of features with solar power output is determined across different months of the year. The simulation results show that the proposed LSTM-based online learning model achieves a mean absolute percentage error of less than 1%. The results demonstrate that incorporating solar transition period information significantly enhances the forecasting accuracy of the proposed model. Online learning exhibits higher performance than batch learning in terms of forecasting accuracy.
- 8 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. PP, 1
Open AccessLSTM-Based Online Learning for Real-Time Solar Power Forecasting
Accurate solar power forecasting is essential to mitigate the variability and uncertainty of solar power generation and integrate renewable energy sources into power grids. A multi-step Long Short Term Memory (LSTM)-based model is proposed for solar power generation forecasting using online learning. To capture daily fluctuations in solar energy patterns, the model employs a dynamic sunrise–sunset data window operation, ensuring that the relevant transitions in solar irradiance are accurately represented in the training set. For seasonality analysis, the variance of the correlation of features with solar power output is determined across different months of the year. The simulation results show that the proposed LSTM-based online learning model achieves a mean absolute percentage error of less than 1%. The results demonstrate that incorporating solar transition period information significantly enhances the forecasting accuracy of the proposed model. Online learning exhibits higher performance than batch learning in terms of forecasting accuracy.
- Open Access
Wavelet-Enhanced PaDiM for Industrial Anomaly Detection
Anomaly detection and localization in industrial images are essential for automated quality inspection. PaDiM, a prominent method, models the distribution of normal image features extracted by pre-trained Convolutional Neural Networks (CNNs) but reduces dimensionality through random channel selection, potentially discarding structured information. We propose Wavelet-Enhanced PaDiM (WE-PaDiM), which integrates Discrete Wavelet Transform (DWT) analysis with multi-layer CNN features in a structured manner. WE-PaDiM applies 2D DWT to feature maps from multiple backbone layers, selects specific frequency subbands (e.g., LL, LH, HL), spatially aligns them, and concatenates them channel-wise before modeling with PaDiM's multivariate Gaussian framework. This DWT-before-concatenation strategy provides a principled method for feature selection based on frequency content relevant to anomalies, leveraging multi-scale wavelet information as an alternative to random selection. We evaluate WE-PaDiM on the challenging MVTec AD dataset with multiple backbones (ResNet-18 and EfficientNet B0-B6). The method achieves strong performance in anomaly detection and localization, yielding average results of 99.32% Image-AUC and 92.10% Pixel-AUC across 15 categories with per-class optimized configurations. Our analysis shows that wavelet choices affect performance trade-offs: simpler wavelets (e.g., Haar) with detail subbands (HL or LH/HL/HH) often enhance localization, while approximation bands (LL) improve image-level detection. WE-PaDiM thus offers a competitive and interpretable alternative to random feature selection in PaDiM, achieving robust results suitable for industrial inspection with comparable efficiency.
Open AccessWavelet-Enhanced PaDiM for Industrial Anomaly Detection
Anomaly detection and localization in industrial images are essential for automated quality inspection. PaDiM, a prominent method, models the distribution of normal image features extracted by pre-trained Convolutional Neural Networks (CNNs) but reduces dimensionality through random channel selection, potentially discarding structured information. We propose Wavelet-Enhanced PaDiM (WE-PaDiM), which integrates Discrete Wavelet Transform (DWT) analysis with multi-layer CNN features in a structured manner. WE-PaDiM applies 2D DWT to feature maps from multiple backbone layers, selects specific frequency subbands (e.g., LL, LH, HL), spatially aligns them, and concatenates them channel-wise before modeling with PaDiM's multivariate Gaussian framework. This DWT-before-concatenation strategy provides a principled method for feature selection based on frequency content relevant to anomalies, leveraging multi-scale wavelet information as an alternative to random selection. We evaluate WE-PaDiM on the challenging MVTec AD dataset with multiple backbones (ResNet-18 and EfficientNet B0-B6). The method achieves strong performance in anomaly detection and localization, yielding average results of 99.32% Image-AUC and 92.10% Pixel-AUC across 15 categories with per-class optimized configurations. Our analysis shows that wavelet choices affect performance trade-offs: simpler wavelets (e.g., Haar) with detail subbands (HL or LH/HL/HH) often enhance localization, while approximation bands (LL) improve image-level detection. WE-PaDiM thus offers a competitive and interpretable alternative to random feature selection in PaDiM, achieving robust results suitable for industrial inspection with comparable efficiency.