Search Results - "Cha, Sungmin"

Toward a unified framework for interpreting machine-learning models in neuroimaging by Kohoutová, Lada, Heo, Juyeon, Cha, Sungmin, Lee, Sungwoo, Moon, Taesup, Wager, Tor D., Woo, Choong-Wan

“…Machine learning is a powerful tool for creating computational models relating brain function to behavior, and its use is becoming widespread in neuroscience…”
Get full text

Applying Low-Impact Development Techniques for Improved Water Management in Urban Areas by Kim, Jaemoon, Park, Jaerock, Cha, Sungmin, Kwon, Soonchul

“…Worldwide, the increase in impervious surfaces due to urbanization has led to significant water cycle issues such as groundwater depletion, urban heat islands,…”
Get full text

Observations on K-image Expansion of Image-Mixing Augmentation by Jeong, Joonhyun, Cha, Sungmin, Choi, Jongwon, Yun, Sangdoo, Moon, Taesup, Yoo, Youngjoon

“…Image-mixing augmentations (e.g., Mixup and CutMix), which typically involve mixing two images, have become the de-facto training techniques for image…”
Get full text

Fully Convolutional Pixel Adaptive Image Denoiser by Cha, Sungmin, Moon, Taesup

“…We propose a new image denoising algorithm, dubbed as Fully Convolutional Adaptive Image DEnoiser (FC-AIDE), that can learn from an offline supervised training…”
Get full text

Neural Adaptive Image Denoiser by Cha, Sungmin, Moon, Taesup

“…We propose a novel neural network-based adaptive image denoiser, dubbased as Neural AIDE. Unlike other neural network-based denoisers, which typically apply…”
Get full text

Enhanced Electrochemical Stability of a Zwitterionic-Polymer-Functionalized Electrode for Capacitive Deionization by Jung, Youngsuk, Yang, Yooseong, Kim, Taeyoon, Shin, Hyun Suk, Hong, Sunghoon, Cha, Sungmin, Kwon, Soonchul

“…In capacitive deionization, the salt-adsorption capacity of the electrode is critical for the efficient softening of brackish water. To improve the…”
Get full text

NCIS: Neural Contextual Iterative Smoothing for Purifying Adversarial Perturbations by Cha, Sungmin, Ko, Naeun, Choi, Heewoong, Yoo, Youngjoon, Moon, Taesup

“…We propose a novel and effective purification-based adversarial defense method against pre-processor blind white-and black-box attacks, without requiring any…”
Get full text

Rebalancing Batch Normalization for Exemplar-Based Class-Incremental Learning by Cha, Sungmin, Cho, Sungjun, Hwang, Dasol, Hong, Sunwon, Lee, Moontae, Moon, Taesup

“…Batch Normalization (BN) and its variants has been extensively studied for neural nets in various computer vision tasks, but relatively little work has been…”
Get full text

FBI-Denoiser: Fast Blind Image Denoiser for Poisson-Gaussian Noise by Byun, Jaeseok, Cha, Sungmin, Moon, Taesup

“…We consider the challenging blind denoising problem for Poisson-Gaussian noise, in which no additional information about clean images or noise level parameters…”
Get full text

Hyperparameters in Continual Learning: A Reality Check by Cha, Sungmin, Cho, Kyunghyun

“…Continual learning (CL) aims to train a model on a sequence of tasks (i.e., a CL scenario) while balancing the trade-off between plasticity (learning new…”
Get full text

Towards Realistic Incremental Scenario in Class Incremental Semantic Segmentation by Kwak, Jihwan, Cha, Sungmin, Moon, Taesup

“…This paper addresses the unrealistic aspect of the commonly adopted Continuous Incremental Semantic Segmentation (CISS) scenario, termed overlapped. We point…”
Get full text

Regularizing with Pseudo-Negatives for Continual Self-Supervised Learning by Cha, Sungmin, Cho, Kyunghyun, Moon, Taesup

“…We introduce a novel Pseudo-Negative Regularization (PNR) framework for effective continual self-supervised learning (CSSL). Our PNR leverages pseudo-negatives…”
Get full text

Towards Robust and Cost-Efficient Knowledge Unlearning for Large Language Models by Cha, Sungmin, Cho, Sungjun, Hwang, Dasol, Lee, Moontae

“…Large Language Models (LLMs) have demonstrated strong reasoning and memorization capabilities via pretraining on massive textual corpora. However, this poses…”
Get full text

Fully Convolutional Pixel Adaptive Image Denoiser by Cha, Sungmin, Moon, Taesup

“…We propose a new image denoising algorithm, dubbed as Fully Convolutional Adaptive Image DEnoiser (FC-AIDE), that can learn from an offline supervised training…”
Get full text

FBI-Denoiser: Fast Blind Image Denoiser for Poisson-Gaussian Noise by Byun, Jaeseok, Cha, Sungmin, Moon, Taesup

“…We consider the challenging blind denoising problem for Poisson-Gaussian noise, in which no additional information about clean images or noise level parameters…”
Get full text

Salience-Based Adaptive Masking: Revisiting Token Dynamics for Enhanced Pre-training by Choi, Hyesong, Park, Hyejin, Yi, Kwang Moo, Cha, Sungmin, Min, Dongbo

“…In this paper, we introduce Saliency-Based Adaptive Masking (SBAM), a novel and cost-effective approach that significantly enhances the pre-training…”
Get full text

Neural Affine Grayscale Image Denoising by Cha, Sungmin, Moon, Taesup

“…We propose a new grayscale image denoiser, dubbed as Neural Affine Image Denoiser (Neural AIDE), which utilizes neural network in a novel way. Unlike other…”
Get full text

Supervised Neural Discrete Universal Denoiser for Adaptive Denoising by Cha, Sungmin, Min, Seonwoo, Yoon, Sungroh, Moon, Taesup

“…We improve the recently developed Neural DUDE, a neural network-based adaptive discrete denoiser, by combining it with the supervised learning framework…”
Get full text

Interpretation of seasonal water quality variation in the Yeongsan Reservoir, Korea using multivariate statistical analyses by Cho, Kyung Hwa, Park, Yongeun, Kang, Joo-Hyon, Ki, Seo Jin, Cha, Sungmin, Lee, Seung Won, Kim, Joon Ha

“…The Yeongsan (YS) Reservoir is an estuarine reservoir which provides surrounding areas with public goods, such as water supply for agricultural and industrial…”
Get full text

SSUL: Semantic Segmentation with Unknown Label for Exemplar-based Class-Incremental Learning by Cha, Sungmin, Kim, Beomyoung, Yoo, Youngjoon, Moon, Taesup

“…This paper introduces a solid state-of-the-art baseline for a class-incremental semantic segmentation (CISS) problem. While the recent CISS algorithms utilize…”
Get full text