Search Results for author: Myungjoo Kang

Found 24 papers, 11 papers with code

Feature-aligned N-BEATS with Sinkhorn divergence

1 code implementation24 May 2023 Myeongho Jeon, Myungjoo Kang, Joonhun Lee, Kyunghyun Park

In this study, we propose Feature-aligned N-BEATS as a domain generalization model for univariate time series forecasting problems.

Domain Generalization Representation Learning +2

Generative Modeling through the Semi-dual Formulation of Unbalanced Optimal Transport

1 code implementation24 May 2023 Jaemoo Choi, Jaewoong Choi, Myungjoo Kang

In this paper, we propose a novel generative model based on the semi-dual formulation of Unbalanced Optimal Transport (UOT).

AnomalyBERT: Self-Supervised Transformer for Time Series Anomaly Detection using Data Degradation Scheme

1 code implementation8 May 2023 Yungi Jeong, Eunseok Yang, Jung Hyun Ryu, Imseong Park, Myungjoo Kang

Mechanical defects in real situations affect observation values and cause abnormalities in multivariate time series, such as sensor values or network data.

Anomaly Detection Time Series +1

Restoration based Generative Models

1 code implementation20 Feb 2023 Jaemoo Choi, Yesom Park, Myungjoo Kang

Also, we propose a multi-scale training, which improves the performance compared to the diffusion process, by taking advantage of the flexibility of the forward process.

Denoising Image Restoration

Minimal Width for Universal Property of Deep RNN

no code implementations25 Nov 2022 Chang hoon Song, Geonho Hwang, Jun Ho Lee, Myungjoo Kang

In this study, we prove the universality of deep narrow RNNs and show that the upper bound of the minimum width for universality can be independent of the length of the data.

Universal Property of Convolutional Neural Networks

no code implementations18 Nov 2022 Geonho Hwang, Myungjoo Kang

A convolution with padding outputs the data of the same shape as the input data; therefore, it is necessary to prove whether a convolutional neural network composed of convolutions can approximate such a function.

Finding the global semantic representation in GAN through Frechet Mean

no code implementations11 Oct 2022 Jaewoong Choi, Geonho Hwang, Hyunsoo Cho, Myungjoo Kang

This semantic basis represents sample-independent meaningful perturbations that change the same semantic attribute of an image on the entire latent space.

Bounding the Rademacher Complexity of Fourier neural operators

no code implementations12 Sep 2022 TaeYoung Kim, Myungjoo Kang

Using capacity based on these norms, we bound the generalization error of the model.

Self-Knowledge Distillation via Dropout

no code implementations11 Aug 2022 Hyoje Lee, Yeachan Park, Hyun Seo, Myungjoo Kang

In this paper, we propose a simple and effective self-knowledge distillation method using a dropout (SD-Dropout).

Adversarial Robustness Image Classification +4

Analyzing the Latent Space of GAN through Local Dimension Estimation

no code implementations26 May 2022 Jaewoong Choi, Geonho Hwang, Hyunsoo Cho, Myungjoo Kang

In this paper, we approach this problem through a geometric analysis of latent spaces as a manifold.

Disentanglement Image Generation

A Conservative Approach for Unbiased Learning on Unknown Biases

1 code implementation CVPR 2022 Myeongho Jeon, Daekyung Kim, Woochul Lee, Myungjoo Kang, Joonseok Lee

Although convolutional neural networks (CNNs) achieve state-of-the-art in image classification, recent works address their unreliable predictions due to their excessive dependence on biased training data.

Image Classification

Do Not Escape From the Manifold: Discovering the Local Coordinates on the Latent Space of GANs

2 code implementations ICLR 2022 Jaewoong Choi, Junho Lee, Changyeon Yoon, Jung Ho Park, Geonho Hwang, Myungjoo Kang

The global warpage implies that the latent space is not well-aligned globally and therefore the global traversal directions are bound to show limited success on it.

Disentanglement Image Generation +1

High-Frequency aware Perceptual Image Enhancement

no code implementations25 May 2021 Hyungmin Roh, Myungjoo Kang

In this paper, we introduce a novel deep neural network suitable for multi-scale analysis and propose efficient model-agnostic methods that help the network extract information from high-frequency domains to reconstruct clearer images.

Deblurring Denoising +3

OAAE: Adversarial Autoencoders for Novelty Detection in Multi-modal Normality Case via Orthogonalized Latent Space

no code implementations7 Jan 2021 Sungkwon An, Jeonghoon Kim, Myungjoo Kang, Shahbaz Razaei, Xin Liu

Specifically, we employ orthogonal low-rank embedding in the latent space to disentangle the features in the latent space using mutual class information.

Image Reconstruction

MCW-Net: Single Image Deraining with Multi-level Connections and Wide Regional Non-local Blocks

1 code implementation29 Sep 2020 Yeachan Park, Myeongho Jeon, Junho Lee, Myungjoo Kang

Experimental results on both synthetic and real-world rainy datasets demonstrate that the proposed model significantly outperforms existing state-of-the-art models.

Single Image Deraining

Discond-VAE: Disentangling Continuous Factors from the Discrete

no code implementations17 Sep 2020 Jaewoong Choi, Geonho Hwang, Myungjoo Kang

To represent these generative factors of data, we introduce two sets of continuous latent variables, private variable and public variable.


Blur Invariant Kernel-Adaptive Network for Single Image Blind deblurring

no code implementations9 Jul 2020 Sungkwon An, Hyungmin Roh, Myungjoo Kang

Subsequently, we propose a deblurring network that restores sharp images using the estimated blur kernel.

Image Deblurring Image Restoration +1

Attention routing between capsules

2 code implementations3 Jul 2019 Jaewoong Choi, Hyun Seo, Suii Im, Myungjoo Kang

We replace the dynamic routing and squash activation function of the capsule network with dynamic routing (CapsuleNet) with the attention routing and capsule activation.

Financial series prediction using Attention LSTM

no code implementations28 Feb 2019 Sangyeon Kim, Myungjoo Kang

In recent times, deep learning methods (especially time series analysis) have performed outstandingly for various industrial problems, with better prediction than machine learning methods.

BIG-bench Machine Learning Time Series +1

Cannot find the paper you are looking for? You can Submit a new open access paper.