Search Results - "Bae, Juhan"

Semi-online video stabilization using probabilistic keyframe update and inter-keyframe motion smoothing by Juhan Bae, Youngbae Hwang, Jongwoo Lim

“…In this paper, we propose a video stabilization method that takes advantages of both online and offline video stabilization methods in a semi-online framework…”
Get full text

Background subtraction using edge cues and color difference for stabilized CMOS images by Bae, Juhan, Hwang, Youngbae, Choi, Byeongho

“…Video stabilization followed by background subtraction using CMOS sensor shows erroneous foregrounds. In this paper, we present a background subtraction method…”
Get full text

Quantitative 3D Flow Visualization of Conventional Purge Flow Within a Front Opening Unified Pod (FOUP) by Lee, Sung-Gwang, Bae, Juhan, Choi, Hoomi, Jeong, Jaein, Kim, Youngjeong, Hwang, Wontae

“…The front opening unified pod (FOUP) is a carrier that transports multiple wafers as it moves between numerous processing facilities. It is inevitably exposed…”
Get full text

Robust visual tracking through deep learning-based confidence evaluation by Au Euntae Hong, Au Juhan Bae, Au Jongwoo Lim

“…In this paper, we propose an object tracking method through deep learning-based confidence evaluation, aiming at correctly updating an object template and…”
Get full text

Fast 6DOF Pose Estimation with Synthetic Textureless CAD Model for Mobile Applications by Chen, Bowen, Bae, Juhan, Mukherjee, Dibyendu

“…Performance of 6DoF pose estimation techniques from RGB/RGBD images has improved significantly with sophisticated deep learning frameworks. These frameworks…”
Get full text

Delta-STN: Efficient Bilevel Optimization for Neural Networks using Structured Response Jacobians by Bae, Juhan, Grosse, Roger

“…Hyperparameter optimization of neural networks can be elegantly formulated as a bilevel optimization problem. While research on bilevel optimization of neural…”
Get full text

Training Data Attribution via Approximate Unrolled Differentiation by Bae, Juhan, Lin, Wu, Lorraine, Jonathan, Grosse, Roger

“…Many training data attribution (TDA) methods aim to estimate how a model's behavior would change if one or more data points were removed from the training set…”
Get full text

Efficient Parametric Approximations of Neural Network Function Space Distance by Dhawan, Nikita, Huang, Sicong, Bae, Juhan, Grosse, Roger

“…It is often useful to compactly summarize important properties of model parameters and training data so that they can be used later without storing and/or…”
Get full text

Influence Functions for Scalable Data Attribution in Diffusion Models by Mlodozeniec, Bruno, Eschenhagen, Runa, Bae, Juhan, Immer, Alexander, Krueger, David, Turner, Richard

“…Diffusion models have led to significant advancements in generative modelling. Yet their widespread adoption poses challenges regarding data attribution and…”
Get full text

If Influence Functions are the Answer, Then What is the Question? by Bae, Juhan, Ng, Nathan, Lo, Alston, Ghassemi, Marzyeh, Grosse, Roger

“…Influence functions efficiently estimate the effect of removing a single training data point on a model's learned parameters. While influence estimates align…”
Get full text

Can We Remove the Square-Root in Adaptive Gradient Methods? A Second-Order Perspective by Lin, Wu, Dangel, Felix, Eschenhagen, Runa, Bae, Juhan, Turner, Richard E, Makhzani, Alireza

“…Adaptive gradient optimizers like Adam(W) are the default training algorithms for many deep learning architectures, such as transformers. Their diagonal…”
Get full text

Using Large Language Models for Hyperparameter Optimization by Zhang, Michael R, Desai, Nishkrit, Bae, Juhan, Lorraine, Jonathan, Ba, Jimmy

“…This paper explores the use of foundational large language models (LLMs) in hyperparameter optimization (HPO). Hyperparameters are critical in determining the…”
Get full text

Amortized Proximal Optimization by Bae, Juhan, Vicol, Paul, HaoChen, Jeff Z, Grosse, Roger

“…We propose a framework for online meta-optimization of parameters that govern optimization, called Amortized Proximal Optimization (APO). We first interpret…”
Get full text

Procedural Knowledge in Pretraining Drives Reasoning in Large Language Models by Ruis, Laura, Mozes, Maximilian, Bae, Juhan, Kamalakara, Siddhartha Rao, Talupuru, Dwarak, Locatelli, Acyr, Kirk, Robert, Rocktäschel, Tim, Grefenstette, Edward, Bartolo, Max

“…The capabilities and limitations of Large Language Models have been sketched out in great detail in recent years, providing an intriguing yet conflicting…”
Get full text

Multi-Rate VAE: Train Once, Get the Full Rate-Distortion Curve by Bae, Juhan, Zhang, Michael R, Ruan, Michael, Wang, Eric, Hasegawa, So, Ba, Jimmy, Grosse, Roger

“…Variational autoencoders (VAEs) are powerful tools for learning latent representations of data used in a wide range of applications. In practice, VAEs usually…”
Get full text

Study on HDR/WCG Service Model for UHD Service by Bae, Juhan, Lim, Jeongyeon, Jung, So Ki

“…In recent years, as people's interest in high-quality media services and technology development have increased, not only methods producing content supporting…”
Get full text

Eigenvalue Corrected Noisy Natural Gradient by Bae, Juhan, Zhang, Guodong, Grosse, Roger

“…Variational Bayesian neural networks combine the flexibility of deep learning with Bayesian uncertainty estimation. However, inference procedures for flexible…”
Get full text

Learnable Pooling Methods for Video Classification by Kmiec, Sebastian, Bae, Juhan, An, Ruijian

“…We introduce modifications to state-of-the-art approaches to aggregating local video descriptors by using attention mechanisms and function approximations…”
Get full text

What is Your Data Worth to GPT? LLM-Scale Data Valuation with Influence Functions by Choe, Sang Keun, Ahn, Hwijeen, Bae, Juhan, Zhao, Kewen, Kang, Minsoo, Chung, Youngseog, Pratapa, Adithya, Neiswanger, Willie, Strubell, Emma, Mitamura, Teruko, Schneider, Jeff, Hovy, Eduard, Grosse, Roger, Xing, Eric

“…Large language models (LLMs) are trained on a vast amount of human-written data, but data providers often remain uncredited. In response to this issue, data…”
Get full text

Analyzing Monotonic Linear Interpolation in Neural Network Loss Landscapes by Lucas, James, Bae, Juhan, Zhang, Michael R, Fort, Stanislav, Zemel, Richard, Grosse, Roger

“…Linear interpolation between initial neural network parameters and converged parameters after training with stochastic gradient descent (SGD) typically leads…”
Get full text