arXiv subjects

Kilho Son

Publications and source records attributed to Kilho Son.

6 recordsLinked to original sources

arXiv2026

RCEM: Robust Conversational Search EMbedder in Distributional Shift

We propose RCEM, a Robust Conversational search EMbedder that is additionally equipped with LLM's query reformulation capability without losing base model's generalization. Unlike prior conversational dense retrieval approaches that learn direct conversation-to-passage matching, RCEM aligns conversations, prepended by special token, to LLM-rewritten queries, while preserving the original embedding space. The unchanged embedding space automatically maps the rewritten-query to the relevant passages. As a result, RCEM (1) reduces overfitting by simplifying the alignment task from long passages to shorter rewritten queries, (2) eliminates the need for conversation-to-passage relevance labels for training, and (3) maintains its original embedding space that allows conversational queries against indexes built by original embedder without rebuilding them. Extensive experiments show that RCEM consistently outperforms prior approaches, achieving up to 30% improvement under distributional shift.

cs.CL↗

arXiv2025

Triaxial Asymmetry Driven Rotational Dynamics and Lateral Equilibrium Position in Inertial Flow

The growing use of triaxial particles in microfluidic, microrobotic, and biological systems makes it essential to understand how their rotational dynamics couples with lateral migration in microscale flows. Our experiments in inertial Poiseuille flow reveal that geometric asymmetry in triaxial, multifaceted disks governs their orientation, migration, and rotational period, distinguishing them from classical axisymmetric objects. We identified a Reynolds- and geometry-dependent shift in preferred rotational orientation, arising from the Dzhanibekov effect, with transition modes determined by the particle's principal-axis configuration. We quantified a scalar offset from Jeffery's orbit prediction and introduced a fitting parameter that generalizes the Jeffery equation to include moment-of-inertia effects on rotational dynamics. Finally, we report the diameter of gyration as a predictor of the lateral equilibrium position of inertially focused triaxial particles. Our results link particle asymmetry to migration and rotation in flow, expanding our understanding of particle dynamics.

physics.flu-dyn↗

arXiv2023

TextManiA: Enriching Visual Feature by Text-driven Manifold Augmentation

We propose TextManiA, a text-driven manifold augmentation method that semantically enriches visual feature spaces, regardless of class distribution. TextManiA augments visual data with intra-class semantic perturbation by exploiting easy-to-understand visually mimetic words, i.e., attributes. This work is built on an interesting hypothesis that general language models, e.g., BERT and GPT, encompass visual information to some extent, even without training on visual training data. Given the hypothesis, TextManiA transfers pre-trained text representation obtained from a well-established large language encoder to a target visual feature space being learned. Our extensive analysis hints that the language encoder indeed encompasses visual information at least useful to augment visual representation. Our experiments demonstrate that TextManiA is particularly powerful in scarce samples with class imbalance as well as even distribution. We also show compatibility with the label mix-based approaches in evenly distributed scarce data.

cs.CV↗

arXiv2023

Learning Customized Visual Models with Retrieval-Augmented Knowledge

Image-text contrastive learning models such as CLIP have demonstrated strong task transfer ability. The high generality and usability of these visual models is achieved via a web-scale data collection process to ensure broad concept coverage, followed by expensive pre-training to feed all the knowledge into model weights. Alternatively, we propose REACT, REtrieval-Augmented CusTomization, a framework to acquire the relevant web knowledge to build customized visual models for target domains. We retrieve the most relevant image-text pairs (~3% of CLIP pre-training data) from the web-scale database as external knowledge, and propose to customize the model by only training new modualized blocks while freezing all the original weights. The effectiveness of REACT is demonstrated via extensive experiments on classification, retrieval, detection and segmentation tasks, including zero, few, and full-shot settings. Particularly, on the zero-shot classification task, compared with CLIP, it achieves up to 5.4% improvement on ImageNet and 3.7% on the ELEVATER benchmark (20 datasets).

cs.CV↗

arXiv2019

Bootstrapping Deep Neural Networks from Approximate Image Processing Pipelines

Complex image processing and computer vision systems often consist of a processing pipeline of functional modules. We intend to replace parts or all of a target pipeline with deep neural networks to achieve benefits such as increased accuracy or reduced computational requirement. To acquire a large amount of labeled data necessary to train the deep neural network, we propose a workflow that leverages the target pipeline to create a significantly larger labeled training set automatically, without prior domain knowledge of the target pipeline. We show experimentally that despite the noise introduced by automated labeling and only using a very small initially labeled data set, the trained deep neural networks can achieve similar or even better performance than the components they replace, while in some cases also reducing computational requirements.

cs.CV↗

arXiv2016

Learning to Remove Multipath Distortions in Time-of-Flight Range Images for a Robotic Arm Setup

Range images captured by Time-of-Flight (ToF) cameras are corrupted with multipath distortions due to interaction between modulated light signals and scenes. The interaction is often complicated, which makes a model-based solution elusive. We propose a learning-based approach for removing the multipath distortions for a ToF camera in a robotic arm setup. Our approach is based on deep learning. We use the robotic arm to automatically collect a large amount of ToF range images containing various multipath distortions. The training images are automatically labeled by leveraging a high precision structured light sensor available only in the training time. In the test time, we apply the learned model to remove the multipath distortions. This allows our robotic arm setup to enjoy the speed and compact form of the ToF camera without compromising with its range measurement errors. We conduct extensive experimental validations and compare the proposed method to several baseline algorithms. The experiment results show that our method achieves 55% error reduction in range estimation and largely outperforms the baseline algorithms.

cs.CV↗