arXiv subjects

Minwoo Park

Publications and source records attributed to Minwoo Park.

12 recordsLinked to original sources

arXiv2026

Uncertainty-Aware Structure-Property Mapping of Spinodoid Metamaterials via Heteroscedastic Gaussian Process Regression

Spinodoid metamaterials offer a broad, tunable design space for anisotropic mechanical properties, yet their structure-property relationships are commonly treated as representative mappings from cone-angle descriptors to single effective stiffness values. This deterministic view overlooks the stochastic nature of Gaussian random field (GRF)-based topology generation, where identical cone-angle descriptors can produce different morphology realizations and property scatter. Here, we present an uncertainty-aware structure-property mapping framework that reinterprets cone-angle descriptors as stochastic descriptors associated with input-dependent property distributions. Using heteroscedastic Gaussian process regression (GPR), the framework infers input-dependent predictive uncertainty from sparse one-realization-per-point data without requiring empirical variance labels at every design point. The results show that stiffness scatter differs across tensor components according to each component's mechanically active directions, and that parameter sets yielding identical mean stiffness can carry different aleatoric uncertainty. Applying this uncertainty to reliability-based design optimization (RBDO), we show that a deterministic optimum is highly susceptible to constraint violation once morphology-induced variability is considered, and that a homoscedastic RBDO formulation fails to meet the prescribed reliability target - only the heteroscedastic formulation satisfies the reliability target under the heteroscedastic uncertainty evaluation. This establishes uncertainty-aware surrogate modeling as essential for reliability-aware inverse design of spinodoid metamaterials; extending the framework to nonlinear responses remains for future work.

physics.comp-ph↗

arXiv2026

Morphology-, Noise-, and Resolution-Robust Ultrasound Elasticity Imaging with Fourier Neural Operators

Ultrasound-based elasticity imaging is a non-invasive technique for estimating tissue stiffness fields from displacement fields obtained by comparing ultrasound signals before and after compression. While recent deep learning approaches have enabled faster and more accurate elasticity estimation compared to traditional methods, several challenges remain for clinical translation. In this study, we employ finite element simulations of free-hand palpation to investigate the applicability of the Fourier neural operator (FNO). Four practical scenarios were investigated: (1) prediction across diverse lesion morphologies, (2) generalization to cases with lesion counts differing from those in the training data, (3) robustness to noise in measured displacement fields, and (4) resilience to variations in ultrasound device resolution. Across these tasks, FNO consistently outperformed baseline models such as U-Net and DeepONet in predictive accuracy and generalization, while maintaining robustness under noise and resolution changes. Validated through in silico simulations, these findings demonstrate the potential of FNO as a framework that could facilitate translation of elasticity imaging toward clinical practice.

physics.med-ph↗

arXiv2025

Symmetry-Aware Steering of Equivariant Diffusion Policies: Benefits and Limits

Equivariant diffusion policies (EDPs) combine the generative expressivity of diffusion models with the strong generalization and sample efficiency afforded by geometric symmetries. While steering these policies with reinforcement learning (RL) offers a promising mechanism for fine-tuning beyond demonstration data, directly applying standard (non-equivariant) RL can be sample-inefficient and unstable, as it ignores the symmetries that EDPs are designed to exploit. In this paper, we theoretically establish that the diffusion process of an EDP is equivariant, which in turn induces a group-invariant latent-noise MDP that is well-suited for equivariant diffusion steering. Building on this theory, we introduce a principled symmetry-aware steering framework and compare standard, equivariant, and approximately equivariant RL strategies through comprehensive experiments across tasks with varying degrees of symmetry. While we identify the practical boundaries of strict equivariance under symmetry breaking, we show that exploiting symmetry during the steering process yields substantial benefits-enhancing sample efficiency, preventing value divergence, and achieving strong policy improvements even when EDPs are trained from extremely limited demonstrations.

cs.LG↗

arXiv2025

Scale-invariant and View-relational Representation Learning for Full Surround Monocular Depth

Recent foundation models demonstrate strong generalization capabilities in monocular depth estimation. However, directly applying these models to Full Surround Monocular Depth Estimation (FSMDE) presents two major challenges: (1) high computational cost, which limits real-time performance, and (2) difficulty in estimating metric-scale depth, as these models are typically trained to predict only relative depth. To address these limitations, we propose a novel knowledge distillation strategy that transfers robust depth knowledge from a foundation model to a lightweight FSMDE network. Our approach leverages a hybrid regression framework combining the knowledge distillation scheme--traditionally used in classification--with a depth binning module to enhance scale consistency. Specifically, we introduce a cross-interaction knowledge distillation scheme that distills the scale-invariant depth bin probabilities of a foundation model into the student network while guiding it to infer metric-scale depth bin centers from ground-truth depth. Furthermore, we propose view-relational knowledge distillation, which encodes structural relationships among adjacent camera views and transfers them to enhance cross-view depth consistency. Experiments on DDAD and nuScenes demonstrate the effectiveness of our method compared to conventional supervised methods and existing knowledge distillation approaches. Moreover, our method achieves a favorable trade-off between performance and efficiency, meeting real-time requirements.

cs.CV↗

arXiv2025

Partially Equivariant Reinforcement Learning in Symmetry-Breaking Environments

Group symmetries provide a powerful inductive bias for reinforcement learning (RL), enabling efficient generalization across symmetric states and actions via group-invariant Markov Decision Processes (MDPs). However, real-world environments almost never realize fully group-invariant MDPs; dynamics, actuation limits, and reward design usually break symmetries, often only locally. Under group-invariant Bellman backups for such cases, local symmetry-breaking introduces errors that propagate across the entire state-action space, resulting in global value estimation errors. To address this, we introduce Partially group-Invariant MDP (PI-MDP), which selectively applies group-invariant or standard Bellman backups depending on where symmetry holds. This framework mitigates error propagation from locally broken symmetries while maintaining the benefits of equivariance, thereby enhancing sample efficiency and generalizability. Building on this framework, we present practical RL algorithms -- Partially Equivariant (PE)-DQN for discrete control and PE-SAC for continuous control -- that combine the benefits of equivariance with robustness to symmetry-breaking. Experiments across Grid-World, locomotion, and manipulation benchmarks demonstrate that PE-DQN and PE-SAC significantly outperform baselines, highlighting the importance of selective symmetry exploitation for robust and sample-efficient RL. Project page: https://pranaboy72.github.io/perl_page/

cs.LG↗

arXiv2025

Data-driven Design of Isotropic and High-Stiffness TPMS-based Amorphousness-Induced Architected Material (TAAM)

For their excellent stiffness-to-weight characteristics, triply periodic minimal surfaces (TPMS) are widely adopted in architected materials. However, their geometric regularity often leads to elastic anisotropy, limiting their effectiveness under complex loading. To address this, we propose TPMS-based amorphousness-induced architected materials (TAAMs), which incorporate controllable geometric disorder as a tunable design variable. This concept of designable amorphousness broadens the geometric design space, enabling the simultaneous optimization of stiffness and isotropy. A data-driven framework integrating computational homogenization with multi-objective Bayesian optimization is employed to discover high-performance TAAMs. Selected designs were fabricated using fused deposition modeling and validated through uniaxial compression tests, showing strong agreement with numerical predictions. Compared to conventional TPMS, TAAMs exhibit significantly improved elastic isotropy while maintaining high stiffness across various relative densities. This approach offers a robust and scalable pathway for developing architected materials tailored to applications requiring isotropic performance, such as biomedical implants, protective systems, and aerospace components.

physics.comp-ph↗

arXiv2025

Effects of Homogeneous Doping on Electron-Phonon Coupling in SrTiO3

Bulk n-type SrTiO3 (STO) has long been known to possess a superconducting ground state at an exceptionally dilute carrier density. This has raised questions about the applicability of the BCS-Eliashberg paradigm with its underlying adiabatic assumption. However, recent experimental reports have set the pairing gap to the critical temperature (Tc) ratio at the BCS value for superconductivity in Nb-doped STO, even though the adiabaticity condition the BCS pairing requires is satisfied over the entire superconducting dome only by the lowest branch of optical phonons. In spite of the strong implications these reports have on specifying the pairing glue, they have not proved sufficient in explaining the magnitude of the optimal doping. This motivated us to apply density functional theory to Nb-doped STO to analyze how the phonon band structures and the electron-phonon coupling evolve with doping. To describe the very low doping concentration, we tuned the homogeneous background charge, from which we obtained a first-principles result on the doping-dependent phonon frequency that is in good agreement with experimental data for Nb-doped STO. Using the EPW code, we obtain the doping-dependent phonon dispersion and the electron-phonon coupling strength. Within the framework of our calculation, we found that the electron-phonon coupling forms a dome in a doping range lower than the experimentally observed superconducting dome of the Nb-doped STO. Additionally, we examined the doping dependence of both the orbital angular momentum quenching in the electron-phonon coupling and the phonon displacement correlation length and found the former to have a strong correlation with our electron-phonon coupling in the overdoped region.

cond-mat.supr-con↗

arXiv2024

Augmented Reality based Simulated Data (ARSim) with multi-view consistency for AV perception networks

Detecting a diverse range of objects under various driving scenarios is essential for the effectiveness of autonomous driving systems. However, the real-world data collected often lacks the necessary diversity presenting a long-tail distribution. Although synthetic data has been utilized to overcome this issue by generating virtual scenes, it faces hurdles such as a significant domain gap and the substantial efforts required from 3D artists to create realistic environments. To overcome these challenges, we present ARSim, a fully automated, comprehensive, modular framework designed to enhance real multi-view image data with 3D synthetic objects of interest. The proposed method integrates domain adaptation and randomization strategies to address covariate shift between real and simulated data by inferring essential domain attributes from real data and employing simulation-based randomization for other attributes. We construct a simplified virtual scene using real data and strategically place 3D synthetic assets within it. Illumination is achieved by estimating light distribution from multiple images capturing the surroundings of the vehicle. Camera parameters from real data are employed to render synthetic assets in each frame. The resulting augmented multi-view consistent dataset is used to train a multi-camera perception network for autonomous vehicles. Experimental results on various AV perception tasks demonstrate the superior performance of networks trained on the augmented dataset.

cs.CV↗

arXiv2023

NVAutoNet: Fast and Accurate 360$^{\circ}$ 3D Visual Perception For Self Driving

Achieving robust and real-time 3D perception is fundamental for autonomous vehicles. While most existing 3D perception methods prioritize detection accuracy, they often overlook critical aspects such as computational efficiency, onboard chip deployment friendliness, resilience to sensor mounting deviations, and adaptability to various vehicle types. To address these challenges, we present NVAutoNet: a specialized Bird's-Eye-View (BEV) perception network tailored explicitly for automated vehicles. NVAutoNet takes synchronized camera images as input and predicts 3D signals like obstacles, freespaces, and parking spaces. The core of NVAutoNet's architecture (image and BEV backbones) relies on efficient convolutional networks, optimized for high performance using TensorRT. More importantly, our image-to-BEV transformation employs simple linear layers and BEV look-up tables, ensuring rapid inference speed. Trained on an extensive proprietary dataset, NVAutoNet consistently achieves elevated perception accuracy, operating remarkably at 53 frames per second on the NVIDIA DRIVE Orin SoC. Notably, NVAutoNet demonstrates resilience to sensor mounting deviations arising from diverse car models. Moreover, NVAutoNet excels in adapting to varied vehicle types, facilitated by inexpensive model fine-tuning procedures that expedite compatibility adjustments.

cs.CV↗

arXiv2023

DPPD: Deformable Polar Polygon Object Detection

Regular object detection methods output rectangle bounding boxes, which are unable to accurately describe the actual object shapes. Instance segmentation methods output pixel-level labels, which are computationally expensive for real-time applications. Therefore, a polygon representation is needed to achieve precise shape alignment, while retaining low computation cost. We develop a novel Deformable Polar Polygon Object Detection method (DPPD) to detect objects in polygon shapes. In particular, our network predicts, for each object, a sparse set of flexible vertices to construct the polygon, where each vertex is represented by a pair of angle and distance in the Polar coordinate system. To enable training, both ground truth and predicted polygons are densely resampled to have the same number of vertices with equal-spaced raypoints. The resampling operation is fully differentable, allowing gradient back-propagation. Sparse polygon predicton ensures high-speed runtime inference while dense resampling allows the network to learn object shapes with high precision. The polygon detection head is established on top of an anchor-free and NMS-free network architecture. DPPD has been demonstrated successfully in various object detection tasks for autonomous driving such as traffic-sign, crosswalk, vehicle and pedestrian objects.

cs.CV↗

arXiv2023

HazardNet: Road Debris Detection by Augmentation of Synthetic Models

We present an algorithm to detect unseen road debris using a small set of synthetic models. Early detection of road debris is critical for safe autonomous or assisted driving, yet the development of a robust road debris detection model has not been widely discussed. There are two main challenges to building a road debris detector: first, data collection of road debris is challenging since hazardous objects on the road are rare to encounter in real driving scenarios; second, the variability of road debris is broad, ranging from a very small brick to a large fallen tree. To overcome these challenges, we propose a novel approach to few-shot learning of road debris that uses semantic augmentation and domain randomization to augment real road images with synthetic models. We constrain the problem domain to uncommon objects on the road and allow the deep neural network, HazardNet, to learn the semantic meaning of road debris to eventually detect unseen road debris. Our results demonstrate that HazardNet is able to accurately detect real road debris when only trained on synthetic objects in augmented images.

cs.CV↗

arXiv2016

High-throughput screening of metal-porphyrin-like graphenes for selective capture of carbon dioxide

Nano-materials, such as metal-organic frameworks, have been considered to capture CO$_2$. However, their application has been limited largely because they exhibit poor selectivity for flue gases and low capture capacity under low pressures. We perform a high-throughput screening for selective CO$_2$ capture from flue gases by using first principles thermodynamics. We find that elements with empty d orbitals selectively attract CO$_2$ from gaseous mixtures under low CO$_2$ pressures at 300 K and release it at ~450 K. CO$_2$ binding to elements involves hybridization of the metal d orbitals with the CO$_2$ $π$ orbitals and CO$_2$-transition metal complexes were observed in experiments. This result allows us to perform high-throughput screening to discover novel promising CO$_2$ capture materials with empty d orbitals and predict their capture performance under various conditions. Moreover, these findings provide physical insights into selective CO$_2$ capture and open a new path to explore CO$_2$ capture materials.

cond-mat.mtrl-sci↗