arXiv subjects

Minwook Kim

Publications and source records attributed to Minwook Kim.

6 recordsLinked to original sources

arXiv2026

Nineteen vectors and a certificate of their phase-retrieval injectivity in $\mathbb{C}^6$

We exhibit nineteen explicit vectors in $\mathbb{C}^6$, with Gaussian-integer entries, that do phase retrieval: the intensities $|\langle a_j,x\rangle|^2$, $j=1,\dots,19$, determine every $x\in\mathbb{C}^6$ up to a unimodular scalar. This gives $m{\mathbb{C}}(6)\le 19=4d-5$, one below the generic count $4d-4$; combined with the lower bound $m{\mathbb{C}}(6)\ge 18$ of Wang and Xu, the minimal measurement number in dimension six is 18 or 19. The claim is verified by an exact rational computation, archived and rerunnable: after an elementary reduction, phase retrieval is equivalent to the emptiness of the real zero set of a system of five quadratic equations in five unknowns, and the archived eliminant of that system is a squarefree integer polynomial of degree fourteen with no real roots.

math.CA↗

arXiv2026

Video-rate holographic telepresence via single-shot, reference-free wavefront measurement

We present a reference-free holographic telepresence system that directly captures and replays complex optical wavefronts from a single intensity speckle measurement. Using a pre-characterized geometric phase diffuser, the incident field self-interferes to form a speckle pattern, from which the wavefront is recovered via a speckle-correlation scattering-matrix approach and refined using smoothed amplitude flow with Nesterov acceleration. The reconstructed phase is directly projected onto a spatial light modulator for holographic replay. We demonstrate volumetric refocusing, dynamic three-dimensional reconstruction, and sustained video-rate operation at approximately 28 frames per second with modest communication bandwidth. The results highlight measurement-driven wavefront acquisition as a practical pathway toward compact and physically faithful holographic telepresence.

physics.optics↗

arXiv2025

DepthFocus: Controllable Depth Estimation for See-Through Scenes

Depth in the real world is rarely singular. Transmissive materials create layered ambiguities that confound conventional perception systems. Existing models remain passive; conventional approaches typically estimate static depth maps anchored to the nearest surface, and even recent multi-head extensions suffer from a representational bottleneck due to fixed feature representations. This stands in contrast to human vision, which actively shifts focus to perceive a desired depth. We introduce \textbf{DepthFocus}, a steerable Vision Transformer that redefines stereo depth estimation as condition-aware control. Instead of extracting fixed features, our model dynamically modulates its computation based on a physical reference depth, integrating dual conditional mechanisms to selectively perceive geometry aligned with the desired focus. Leveraging a newly curated large-scale synthetic dataset, \textbf{DepthFocus} achieves state-of-the-art results across all evaluated benchmarks, including both standard single-layer and complex multi-layered scenarios. While maintaining high precision in opaque regions, our approach effectively resolves depth ambiguities in transparent and reflective scenes by selectively reconstructing geometry at a target distance. This capability enables robust, intent-driven perception that significantly outperforms existing multi-layer methods, marking a substantial step toward active 3D perception. \noindent \textbf{Project page}: \href{https://junhong-3dv.github.io/depthfocus-project/}{\textbf{this https URL}}.

cs.CV↗

arXiv2023

CAST: Cluster-Aware Self-Training for Tabular Data via Reliable Confidence

Tabular data is one of the most widely used data modalities, encompassing numerous datasets with substantial amounts of unlabeled data. Despite this prevalence, there is a notable lack of simple and versatile methods for utilizing unlabeled data in the tabular domain, where both gradient-boosting decision trees and neural networks are employed. In this context, self-training has gained attraction due to its simplicity and versatility, yet it is vulnerable to noisy pseudo-labels caused by erroneous confidence. Several solutions have been proposed to handle this problem, but they often compromise the inherent advantages of self-training, resulting in limited applicability in the tabular domain. To address this issue, we explore a novel direction of reliable confidence in self-training contexts and conclude that self-training can be improved by making that the confidence, which represents the value of the pseudo-label, aligns with the cluster assumption. In this regard, we propose Cluster-Aware Self-Training (CAST) for tabular data, which enhances existing self-training algorithms at a negligible cost while maintaining simplicity and versatility. Concretely, CAST calibrates confidence by regularizing the classifier's confidence based on local density for each class in the labeled training data, resulting in lower confidence for pseudo-labels in low-density regions. Extensive empirical evaluations on up to 21 real-world datasets confirm not only the superior performance of CAST but also its robustness in various setups in self-training contexts.

cs.LG↗

arXiv2023

Revisiting Self-Training with Regularized Pseudo-Labeling for Tabular Data

Recent progress in semi- and self-supervised learning has caused a rift in the long-held belief about the need for an enormous amount of labeled data for machine learning and the irrelevancy of unlabeled data. Although it has been successful in various data, there is no dominant semi- and self-supervised learning method that can be generalized for tabular data (i.e. most of the existing methods require appropriate tabular datasets and architectures). In this paper, we revisit self-training which can be applied to any kind of algorithm including the most widely used architecture, gradient boosting decision tree, and introduce curriculum pseudo-labeling (a state-of-the-art pseudo-labeling technique in image) for a tabular domain. Furthermore, existing pseudo-labeling techniques do not assure the cluster assumption when computing confidence scores of pseudo-labels generated from unlabeled data. To overcome this issue, we propose a novel pseudo-labeling approach that regularizes the confidence scores based on the likelihoods of the pseudo-labels so that more reliable pseudo-labels which lie in high density regions can be obtained. We exhaustively validate the superiority of our approaches using various models and tabular datasets.

cs.LG↗

arXiv2022

3D Dual-Fusion: Dual-Domain Dual-Query Camera-LiDAR Fusion for 3D Object Detection

Fusing data from cameras and LiDAR sensors is an essential technique to achieve robust 3D object detection. One key challenge in camera-LiDAR fusion involves mitigating the large domain gap between the two sensors in terms of coordinates and data distribution when fusing their features. In this paper, we propose a novel camera-LiDAR fusion architecture called, 3D Dual-Fusion, which is designed to mitigate the gap between the feature representations of camera and LiDAR data. The proposed method fuses the features of the camera-view and 3D voxel-view domain and models their interactions through deformable attention. We redesign the transformer fusion encoder to aggregate the information from the two domains. Two major changes include 1) dual query-based deformable attention to fuse the dual-domain features interactively and 2) 3D local self-attention to encode the voxel-domain queries prior to dual-query decoding. The results of an experimental evaluation show that the proposed camera-LiDAR fusion architecture achieved competitive performance on the KITTI and nuScenes datasets, with state-of-the-art performances in some 3D object detection benchmarks categories.

cs.CV↗