arXiv subjects

Sungmin Woo

Publications and source records attributed to Sungmin Woo.

At least 19 recordsLinked to original sources

arXiv2026

MoRGS: Efficient Per-Gaussian Motion Reasoning for Streamable Dynamic 3D Scenes

Online reconstruction of dynamic scenes aims to learn from streaming multi-view inputs under low-latency constraints. The fast training and real-time rendering capabilities of 3D Gaussian Splatting have made on-the-fly reconstruction practically feasible, enabling online 4D reconstruction. However, existing online approaches, despite their efficiency and visual quality, fail to learn per-Gaussian motion that reflects true scene dynamics. Without explicit motion cues, appearance and motion are optimized solely under photometric loss, causing per-Gaussian motion to chase pixel residuals rather than true 3D motion. To address this, we propose MoRGS, an efficient online per-Gaussian motion reasoning framework that explicitly models per-Gaussian motion to improve 4D reconstruction quality. Specifically, we leverage optical flow on a sparse set of key views as lightweight motion cues that regularize per-Gaussian motion beyond photometric supervision. To compensate for the sparsity of flow supervision, we learn a per-Gaussian motion offset field that reconciles discrepancies between projected 3D motion and observed flow across views and time. In addition, we introduce a per-Gaussian motion confidence that separates dynamic from static Gaussians and weights Gaussian attribute residual updates, thereby suppressing redundant motion in static regions for better temporal consistency and accelerating the modeling of large motions. Extensive experiments demonstrate that MoRGS achieves state-of-the-art reconstruction quality and motion fidelity among online methods, while maintaining streamable performance.

cs.CV↗

arXiv2026

Revisiting Weakly-Supervised Video Scene Graph Generation via Pair Affinity Learning

Weakly-supervised video scene graph generation (WS-VSGG) aims to parse video content into structured relational triplets without bounding box annotations and with only sparse temporal labeling, significantly reducing annotation costs. Without ground-truth bounding boxes, these methods rely on off-the-shelf detectors to generate object proposals, yet largely overlook a fundamental discrepancy from fullysupervised pipelines. Fully-supervised detectors implicitly filter out noninteractive objects, while off-the-shelf detectors indiscriminately detect all visible objects, overwhelming relation models with noisy pairs.We address this by introducing a learnable pair affinity that estimates the likelihood of interaction between subject-object pairs. Through Pair Affinity Learning and Scoring (PALS), pair affinity is incorporated into inferencetime ranking and further integrated into contextual reasoning through Pair Affinity Modulation (PAM), enabling the model to suppress noninteractive pairs and focus on relationally meaningful ones. To provide cleaner supervision for pair affinity learning, we further propose Relation- Aware Matching (RAM), which leverages vision-language grounding to resolve class-level ambiguity in pseudo-label generation. Extensive experiments on Action Genome demonstrate that our approach consistently yields substantial improvements across different baselines and backbones, achieving state-of-the-art WS-VSGG performance.

cs.CV↗

arXiv2026

Cross Pseudo Labeling For Weakly Supervised Video Anomaly Detection

Weakly supervised video anomaly detection aims to detect anomalies and identify abnormal categories with only video-level labels. We propose CPL-VAD, a dual-branch framework with cross pseudo labeling. The binary anomaly detection branch focuses on snippet-level anomaly localization, while the category classification branch leverages vision-language alignment to recognize abnormal event categories. By exchanging pseudo labels, the two branches transfer complementary strengths, combining temporal precision with semantic discrimination. Experiments on XD-Violence and UCF-Crime demonstrate that CPL-VAD achieves state-of-the-art performance in both anomaly detection and abnormal category classification.

cs.CV↗

arXiv2025

DualFocus: Depth from Focus with Spatio-Focal Dual Variational Constraints

Depth-from-Focus (DFF) enables precise depth estimation by analyzing focus cues across a stack of images captured at varying focal lengths. While recent learning-based approaches have advanced this field, they often struggle in complex scenes with fine textures or abrupt depth changes, where focus cues may become ambiguous or misleading. We present DualFocus, a novel DFF framework that leverages the focal stack's unique gradient patterns induced by focus variation, jointly modeling focus changes over spatial and focal dimensions. Our approach introduces a variational formulation with dual constraints tailored to DFF: spatial constraints exploit gradient pattern changes across focus levels to distinguish true depth edges from texture artifacts, while focal constraints enforce unimodal, monotonic focus probabilities aligned with physical focus behavior. These inductive biases improve robustness and accuracy in challenging regions. Comprehensive experiments on four public datasets demonstrate that DualFocus consistently outperforms state-of-the-art methods in both depth accuracy and perceptual quality.

cs.CV↗

arXiv2024

Improving Unsupervised Video Object Segmentation via Fake Flow Generation

Unsupervised video object segmentation (VOS), also known as video salient object detection, aims to detect the most prominent object in a video at the pixel level. Recently, two-stream approaches that leverage both RGB images and optical flow maps have gained significant attention. However, the limited amount of training data remains a substantial challenge. In this study, we propose a novel data generation method that simulates fake optical flows from single images, thereby creating large-scale training data for stable network learning. Inspired by the observation that optical flow maps are highly dependent on depth maps, we generate fake optical flows by refining and augmenting the estimated depth maps of each image. By incorporating our simulated image-flow pairs, we achieve new state-of-the-art performance on all public benchmark datasets without relying on complex modules. We believe that our data generation method represents a potential breakthrough for future VOS research.

cs.CV↗

arXiv2024

ProDepth: Boosting Self-Supervised Multi-Frame Monocular Depth with Probabilistic Fusion

Self-supervised multi-frame monocular depth estimation relies on the geometric consistency between successive frames under the assumption of a static scene. However, the presence of moving objects in dynamic scenes introduces inevitable inconsistencies, causing misaligned multi-frame feature matching and misleading self-supervision during training. In this paper, we propose a novel framework called ProDepth, which effectively addresses the mismatch problem caused by dynamic objects using a probabilistic approach. We initially deduce the uncertainty associated with static scene assumption by adopting an auxiliary decoder. This decoder analyzes inconsistencies embedded in the cost volume, inferring the probability of areas being dynamic. We then directly rectify the erroneous cost volume for dynamic areas through a Probabilistic Cost Volume Modulation (PCVM) module. Specifically, we derive probability distributions of depth candidates from both single-frame and multi-frame cues, modulating the cost volume by adaptively fusing those distributions based on the inferred uncertainty. Additionally, we present a self-supervision loss reweighting strategy that not only masks out incorrect supervision with high uncertainty but also mitigates the risks in remaining possible dynamic areas in accordance with the probability. Our proposed method excels over state-of-the-art approaches in all metrics on both Cityscapes and KITTI datasets, and demonstrates superior generalization ability on the Waymo Open dataset.

cs.CV↗

arXiv2024

DCE-Qnet: Deep Network Quantification of Dynamic Contrast Enhanced (DCE) MRI

Introduction: Quantification of dynamic contrast-enhanced (DCE)-MRI has the potential to provide valuable clinical information, but robust pharmacokinetic modeling remains a challenge for clinical adoption. Methods: A 7-layer neural network called DCE-Qnet was trained on simulated DCE-MRI signals derived from the Extended Tofts model with the Parker arterial input function. Network training incorporated B1 inhomogeneities to estimate perfusion (Ktrans, vp, ve), tissue T1 relaxation, proton density and bolus arrival time (BAT). The accuracy was tested in a digital phantom in comparison to a conventional nonlinear least-squares fitting (NLSQ). In vivo testing was conducted in 10 healthy subjects. Regions of interest in the cervix and uterine myometrium were used to calculate the inter-subject variability. The clinical utility was demonstrated on a cervical cancer patient. Test-retest experiments were used to assess reproducibility of the parameter maps in the tumor. Results: The DCE-Qnet reconstruction outperformed NLSQ in the phantom. The coefficient of variation (CV) in the healthy cervix varied between 5-51% depending on the parameter. Parameter values in the tumor agreed with previous studies despite differences in methodology. The CV in the tumor varied between 1-47%. Conclusion: The proposed approach provides comprehensive DCE-MRI quantification from a single acquisition. DCE-Qnet eliminates the need for separate T1 scan or BAT processing, leading to a reduction of 10 minutes per scan and more accurate quantification.

physics.med-ph↗

arXiv2024

FIMP: Future Interaction Modeling for Multi-Agent Motion Prediction

Multi-agent motion prediction is a crucial concern in autonomous driving, yet it remains a challenge owing to the ambiguous intentions of dynamic agents and their intricate interactions. Existing studies have attempted to capture interactions between road entities by using the definite data in history timesteps, as future information is not available and involves high uncertainty. However, without sufficient guidance for capturing future states of interacting agents, they frequently produce unrealistic trajectory overlaps. In this work, we propose Future Interaction modeling for Motion Prediction (FIMP), which captures potential future interactions in an end-to-end manner. FIMP adopts a future decoder that implicitly extracts the potential future information in an intermediate feature-level, and identifies the interacting entity pairs through future affinity learning and top-k filtering strategy. Experiments show that our future interaction modeling improves the performance remarkably, leading to superior performance on the Argoverse motion forecasting benchmark.

cs.CV↗

arXiv2023

Leveraging Spatio-Temporal Dependency for Skeleton-Based Action Recognition

Skeleton-based action recognition has attracted considerable attention due to its compact representation of the human body's skeletal sructure. Many recent methods have achieved remarkable performance using graph convolutional networks (GCNs) and convolutional neural networks (CNNs), which extract spatial and temporal features, respectively. Although spatial and temporal dependencies in the human skeleton have been explored separately, spatio-temporal dependency is rarely considered. In this paper, we propose the Spatio-Temporal Curve Network (STC-Net) to effectively leverage the spatio-temporal dependency of the human skeleton. Our proposed network consists of two novel elements: 1) The Spatio-Temporal Curve (STC) module; and 2) Dilated Kernels for Graph Convolution (DK-GC). The STC module dynamically adjusts the receptive field by identifying meaningful node connections between every adjacent frame and generating spatio-temporal curves based on the identified node connections, providing an adaptive spatio-temporal coverage. In addition, we propose DK-GC to consider long-range dependencies, which results in a large receptive field without any additional parameters by applying an extended kernel to the given adjacency matrices of the graph. Our STC-Net combines these two modules and achieves state-of-the-art performance on four skeleton-based action recognition benchmarks.

cs.CV↗

arXiv2023

MKConv: Multidimensional Feature Representation for Point Cloud Analysis

Despite the remarkable success of deep learning, an optimal convolution operation on point clouds remains elusive owing to their irregular data structure. Existing methods mainly focus on designing an effective continuous kernel function that can handle an arbitrary point in continuous space. Various approaches exhibiting high performance have been proposed, but we observe that the standard pointwise feature is represented by 1D channels and can become more informative when its representation involves additional spatial feature dimensions. In this paper, we present Multidimensional Kernel Convolution (MKConv), a novel convolution operator that learns to transform the point feature representation from a vector to a multidimensional matrix. Unlike standard point convolution, MKConv proceeds via two steps. (i) It first activates the spatial dimensions of local feature representation by exploiting multidimensional kernel weights. These spatially expanded features can represent their embedded information through spatial correlation as well as channel correlation in feature space, carrying more detailed local structure information. (ii) Then, discrete convolutions are applied to the multidimensional features which can be regarded as a grid-structured matrix. In this way, we can utilize the discrete convolutions for point cloud data without voxelization that suffers from information loss. Furthermore, we propose a spatial attention module, Multidimensional Local Attention (MLA), to provide comprehensive structure awareness within the local point set by reweighting the spatial feature dimensions. We demonstrate that MKConv has excellent applicability to point cloud processing tasks including object classification, object part segmentation, and scene semantic segmentation with superior results.

cs.CV↗

arXiv2021

Regularization Strategy for Point Cloud via Rigidly Mixed Sample

Data augmentation is an effective regularization strategy to alleviate the overfitting, which is an inherent drawback of the deep neural networks. However, data augmentation is rarely considered for point cloud processing despite many studies proposing various augmentation methods for image data. Actually, regularization is essential for point clouds since lack of generality is more likely to occur in point cloud due to small datasets. This paper proposes a Rigid Subset Mix (RSMix), a novel data augmentation method for point clouds that generates a virtual mixed sample by replacing part of the sample with shape-preserved subsets from another sample. RSMix preserves structural information of the point cloud sample by extracting subsets from each sample without deformation using a neighboring function. The neighboring function was carefully designed considering unique properties of point cloud, unordered structure and non-grid. Experiments verified that RSMix successfully regularized the deep neural networks with remarkable improvement for shape classification. We also analyzed various combinations of data augmentations including RSMix with single and multi-view evaluations, based on abundant ablation studies.

cs.CV↗

arXiv2020

Crystalline symmetry-dependent magnon formation in itinerant ferromagnet SrRuO3

SrRuO3 (SRO) is an itinerant ferromagnet with strong coupling between the charge, spin, and lattice degrees of freedom. This strong coupling suggests that the electronic and magnetic behaviors of SRO are highly susceptible to changes in the lattice distortion. Here we show how the spin interaction and resultant magnon formation change with the modification in the crystallographic orientation. We fabricated SRO epitaxial thin films with (100), (110), and (111) surface orientations, to systematically modulate the spin interaction and spin dimensionality. The reduced spin dimensionality and enhanced exchange interaction in the (111)-oriented SRO thin film significantly suppresses magnon formation. Our study comprehensively demonstrates the facile tunability of magnon formation and spin interaction in correlated oxide thin films.

cond-mat.str-el↗

arXiv2020

PMVOS: Pixel-Level Matching-Based Video Object Segmentation

Semi-supervised video object segmentation (VOS) aims to segment arbitrary target objects in video when the ground truth segmentation mask of the initial frame is provided. Due to this limitation of using prior knowledge about the target object, feature matching, which compares template features representing the target object with input features, is an essential step. Recently, pixel-level matching (PM), which matches every pixel in template features and input features, has been widely used for feature matching because of its high performance. However, despite its effectiveness, the information used to build the template features is limited to the initial and previous frames. We address this issue by proposing a novel method-PM-based video object segmentation (PMVOS)-that constructs strong template features containing the information of all past frames. Furthermore, we apply self-attention to the similarity maps generated from PM to capture global dependencies. On the DAVIS 2016 validation set, we achieve new state-of-the-art performance among real-time methods (> 30 fps), with a J&F score of 85.6%. Performance on the DAVIS 2017 and YouTube-VOS validation sets is also impressive, with J&F scores of 74.0% and 68.2%, respectively.

cs.CV↗

arXiv2020

False Positive Removal for 3D Vehicle Detection with Penetrated Point Classifier

Recently, researchers have been leveraging LiDAR point cloud for higher accuracy in 3D vehicle detection. Most state-of-the-art methods are deep learning based, but are easily affected by the number of points generated on the object. This vulnerability leads to numerous false positive boxes at high recall positions, where objects are occasionally predicted with few points. To address the issue, we introduce Penetrated Point Classifier (PPC) based on the underlying property of LiDAR that points cannot be generated behind vehicles. It determines whether a point exists behind the vehicle of the predicted box, and if does, the box is distinguished as false positive. Our straightforward yet unprecedented approach is evaluated on KITTI dataset and achieved performance improvement of PointRCNN, one of the state-of-the-art methods. The experiment results show that precision at the highest recall position is dramatically increased by 15.46 percentage points and 14.63 percentage points on the moderate and hard difficulty of car class, respectively.

cs.CV↗

arXiv2019

Phase Instability amid Dimensional Crossover in Artificial Oxide Crystal

Artificial crystals synthesized by atomic-scale epitaxy provides the ability to control the dimensions of the quantum phases and associated phase transitions via precise thickness modulation. In particular, reduction in dimensionality via quantized control of atomic layers is a powerful approach to revealing hidden electronic and magnetic phases. Here, we demonstrate a dimensionality-controlled and induced metal-insulator transition (MIT) in atomically designed superlattices by synthesizing a genuine two dimensional (2D) SrRuO3 crystal with highly suppressed charge transfer. The tendency to ferromagnetically align the spins in SrRuO3 layer diminishes in 2D as the interlayer exchange interaction vanishes, accompanying the 2D localization of electrons. Furthermore, electronic and magnetic instabilities in the two SrRuO3 unit cell layers induce a thermally-driven MIT along with a metamagnetic transition.

cond-mat.str-el↗

arXiv2018

Oxygen vacancy-induced structural evolution of SrFeO$_{3-x}$ epitaxial thin film from brownmillerite to perovskite

We investigated SrFeO$_{3-x}$ thin films on a SrTiO$_3$ (001) substrate prepared via pulsed laser epitaxy using an optical spectroscopy technique. The oxygen vacancy level ($x$) was controlled by post-annealing processes at different oxygen partial pressures. We achieved a brownmillerite(BM) structure at $x =$ 0.5 and observed the evolution of the crystal structure from BM into perovskite(PV) as the oxygen concentration increased. We observed the evolution of infrared-active phonons with respect to the oxygen concentration, which was closely related to the structural evolution observed via X-ray diffraction. We identified the phonons using the shell-model calculation. Furthermore, we studied temperature-dependent behaviors of the phonon modes of three representative samples: PV, and two BMs (BM$_{\mathrm{oop}}$ and BM$_{\mathrm{ip}}$) with different orientations of the oxygen vacancy channel. In the BM$_{\mathrm{oop}}$ sample, we observed a phonon mode, which exhibited an unusual red-shift with decreasing temperature; this behavior may have been due to the apical oxygen instability in the FeO$_6$ octahedron. Our results provide important information regarding the ionic conduction mechanism in SrFeO$_{3-x}$ material systems.

cond-mat.mtrl-sci↗

arXiv2018

Enhanced magnetic and thermoelectric properties in epitaxial polycrystalline SrRuO3 thin film

Transition metal oxide thin films show versatile electrical, magnetic, and thermal properties which can be tailored by deliberately introducing macroscopic grain boundaries via polycrystalline solids. In this study, we focus on the modification of the magnetic and thermal transport properties by fabricating single- and polycrystalline epitaxial SrRuO3 thin films using pulsed laser epitaxy. Using epitaxial stabilization technique with atomically flat polycrystalline SrTiO3 substrate, epitaxial polycrystalline SrRuO3 thin film with crystalline quality of each grain comparable to that of single-crystalline counterpart is realized. In particular, alleviated compressive strain near the grain boundaries due to coalescence is evidenced structurally, which induced enhancement of ferromagnetic ordering of the polycrystalline epitaxial thin film. The structural variations associated with the grain boundaries further reduce the thermal conductivity without deteriorating the electronic transport, and lead to enhanced thermoelectric efficiency in the epitaxial polycrystalline thin films, compared with their single-crystalline counterpart.

cond-mat.mtrl-sci↗

arXiv2017

Enhanced electrocatalytic activity via phase transitions in strongly correlated SrRuO3 thin films

Transition metal oxides have been extensively studied and utilized as efficient catalysts. However, the strongly correlated behavior which often results in intriguing emergent phenomena in these materials has been mostly overlooked in understanding the electrochemical activities. Here, we demonstrate a close correlation between the phase transitions and oxygen evolution reaction (OER) in a strongly correlated SrRuO3. By systematically introducing Ru-O vacancies into the single-crystalline SrRuO3 epitaxial thin films, we induced phase transition in crystalline symmetry which resulted in corresponding modification in the electronic structure. The modified electronic structure significantly affect the electrochemical activities, so a 30% decrease in the overpotential for the OER activity was achieved. Our study suggests that a substantial enhancement in the OER activity can be realized even within single material systems, by rational design and engineering of their crystal and electronic structures.

cond-mat.str-el↗