Author name cluster

Dongwon Kim

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

7 papers

2 author rows

ICML Conference 2025 Conference Paper

Federated Learning for Feature Generalization with Convex Constraints

Dongwon Kim
Donghee Kim
Sung Kuk Shyn
Kwangsu Kim

Federated learning (FL) often struggles with generalization due to heterogeneous client data. Local models are prone to overfitting their local data distributions, and even transferable features can be distorted during aggregation. To address these challenges, we propose FedCONST, an approach that adaptively modulates update magnitudes based on the global model’s parameter strength. This prevents over-emphasizing well-learned parameters while reinforcing underdeveloped ones. Specifically, FedCONST employs linear convex constraints to ensure training stability and preserve locally learned generalization capabilities during aggregation. A Gradient Signal-to-Noise Ratio (GSNR) analysis further validates FedCONST’s effectiveness in enhancing feature transferability and robustness. As a result, FedCONST effectively aligns local and global objectives, mitigating overfitting and promoting stronger generalization across diverse FL environments, achieving state-of-the-art performance.

Details

ICML Conference 2025 Conference Paper

One-Step Generalization Ratio Guided Optimization for Domain Generalization

Sumin Cho
Dongwon Kim
Kwangsu Kim

Domain Generalization (DG) aims to train models that generalize to unseen target domains but often overfit to domain-specific features, known as undesired correlations. Gradient-based DG methods typically guide gradients in a dominant direction but often inadvertently reinforce spurious correlations. Recent work has employed dropout to regularize overconfident parameters, but has not explicitly adjusted gradient alignment or ensured balanced parameter updates. We propose GENIE (Generalization-ENhancing Iterative Equalizer), a novel optimizer that leverages the One-Step Generalization Ratio (OSGR) to quantify each parameter’s contribution to loss reduction and assess gradient alignment. By dynamically equalizing OSGR via a preconditioning factor, GENIE prevents a small subset of parameters from dominating optimization, thereby promoting domain-invariant feature learning. Theoretically, GENIE balances convergence contribution and gradient alignment among parameters, achieving higher OSGR while retaining SGD’s convergence rate. Empirically, it outperforms existing optimizers and enhances performance when integrated with various DG and single-DG methods.

Details

NeurIPS Conference 2025 Conference Paper

Partial Correlation Network Estimation by Semismooth Newton Methods

DongWon Kim
Sungdong Lee
Joong-Ho (Johann) Won

We develop a scalable second-order algorithm for a recently proposed $\ell_1$-regularized pseudolikelihood-based partial correlation network estimation framework. While the latter method admits statistical guarantees and is inherently scalable compared to likelihood-based methods such as graphical lasso, the currently available implementations rely only on first-order information and require thousands of iterations to obtain reliable estimates even on high-performance supercomputers. In this paper, we further investigate the inherent scalability of the framework and propose locally and globally convergent semismooth Newton methods. Despite the nonsmoothness of the problem, these second-order algorithms converge at a locally quadratic rate, and require only a few tens of iterations in practice. Each iteration reduces to solving linear systems of small dimensions or linear complementary problems of smaller dimensions, making the computation also suitable for less powerful computing environments. Experiments on both simulated and real-world genomic datasets demonstrate the superior convergence behavior and computational efficiency of the proposed algorithm, which position our method as a promising tool for massive-scale network analysis sought for in, e. g. , modern multi-omics research.

PDF Details

NeurIPS Conference 2024 Conference Paper

Bootstrapping Top-down Information for Self-modulating Slot Attention

DongWon Kim
Seoyeon Kim
Suha Kwak

Object-centric learning (OCL) aims to learn representations of individual objects within visual scenes without manual supervision, facilitating efficient and effective visual reasoning. Traditional OCL methods primarily employ bottom-up approaches that aggregate homogeneous visual features to represent objects. However, in complex visual environments, these methods often fall short due to the heterogeneous nature of visual features within an object. To address this, we propose a novel OCL framework incorporating a top-down pathway. This pathway first bootstraps the semantics of individual objects and then modulates the model to prioritize features relevant to these semantics. By dynamically modulating the model based on its own output, our top-down pathway enhances the representational quality of objects. Our framework achieves state-of-the-art performance across multiple synthetic and real-world object-discovery benchmarks.

PDF Details DOI

ICRA Conference 2020 Conference Paper

Simultaneous Estimations of Joint Angle and Torque in Interactions with Environments using EMG

Dongwon Kim
Kyung Koh
Giovanni Oppizzi
Raziyeh Baghi
Li-Chuan Lo
Chunyang Zhang
Li-Qun Zhang

We develop a decoding technique that estimates both the position and torque of a joint of the limb in interaction with an environment based on activities of the agonist-antagonist pair of muscles using electromyography in real time. The long short-term memory (LSTM) network is employed as the core processor of the proposed technique that is capable of learning time series of a long-time span with varying time lags. A validation that is conducted on the wrist joint shows that the decoding approach provides an agreement of greater than 95% in kinetics (i. e. torque) estimation and an agreement of greater than 85% in kinematics (i. e. angle) estimation, between the actual and estimated variables, during interactions with an environment. Also demonstrated is the fact that the proposed decoding method inherits the strengths of the LSTM network in terms of the capability of learning EMG signals and the corresponding responses with time dependency.

Details

ICRA Conference 2019 Conference Paper

A Simple but Robust Impedance Controller for Series Elastic Actuators

Dongwon Kim
Kyung Koh
Gun-Rae Cho
Li-Qun Zhang

This study presents an impedance controller for series elastic actuators (SEAs), using the singular perturbation (SP) theory and time-delay estimation (TDE) technique. While the SP theory attenuates the requirement for states to be measured, the TDE technique eliminates the requirement for identifying system parameters. Through a numerical analysis and experimental validation, we demonstrate that the proposed controller produces satisfactory tracking performance while-at the same time-pursues wider operational bandwidth and lower driving-point impedance.

Details

IROS Conference 2017 Conference Paper

Impedance control with structural compliance and a sensorless strategy for contact tasks

Dongwon Kim
Sang Hoon Kang
Gwang Min Gu
Maolin Jin

This study proposes an actuator whose design relies on a coordinated approach to control and hardware design: impedance control is supplemented with the introduction of a spring-damper coupler between the actuator and reference (ground). The coupler between the actuator and reference has the effect of further reducing the impedance apparent to the environment. Unlike traditional series elastic actuation, where the deformation of a physical spring coupler introduced between an actuator and link is fed back for control, our coupler is located between actuator and reference and its deformation is not fed back for control. The proposed control further employs time-delay estimation to account for the system dynamics that include both the physical and virtual couplers and to realize accurate and robust control over end-effector position and contact forces. We present a simple method for estimating interaction forces and regulating the contact force without a priori knowledge of the environment. A numerical simulation demonstrates the efficacy of the proposed approach.

Details