arXiv subjects

Koichi Kise

Publications and source records attributed to Koichi Kise.

At least 19 recordsLinked to original sources

arXiv2026

Towards a Joint Khmer Text Recognition and Word Segmentation

Text recognition, or extracting electronic text from document images, has been indispensable for knowledge retrieval tasks, such as retrieval-augmented generation (RAG). For Khmer, extracted text is subject to an extra word segmentation step, as Khmer does not use any visible word delimiters to denote word boundaries. Thus, a recognition-then-segmentation pipeline for Khmer requires two separate sequential models; this is not only error-prone but also adds significant latency for large-scale document processing. This paper proposes a novel joint Khmer text recognition and word segmentation framework in a unified model. The proposed model, using a connectionist-temporal-classification (CTC) decoder for fast, parallel decoding, can be instructed to recognize Khmer text with ($b=1$) and without ($b=0$) word segmentation. Experimental results on different benchmark datasets of different document modalities (document, scene, and handwritten images) show that the proposed model can not only recognize characters in document images but also locate word boundaries, removing the need for an extra word segmentation step in a conventional sequential pipeline.

cs.CV↗

arXiv2026

Empowering Vocabulary Learning Through Teaching AI: Using LLMs as a Student to Perform Learning by Teaching in Vocabulary Acquisition

"Learning by Teaching (LbT)" helps learners deepen their understanding by explaining concepts to others, with questions playing a vital role in identifying knowledge gaps and reinforcing comprehension. However, existing systems for generating such questions often rely on rigid templates and are expensive to build. To overcome these limitations, we developed a system using Large Language Models (LLMs) to create dynamic, contextually relevant questions for LbT. In our English vocabulary learning study, we examined which learner characteristics best leverage the system's benefits. Our results showed improved memory retention over traditional methods at three and seven days of testing, with ten participants. Additionally, we identified traits linked to better learning outcomes, highlighting the potential for tailored approaches. These findings support the development of scalable, cost-effective solutions to enhance LbT methods across various fields.

cs.HC↗

arXiv2026

A Large-Scale Comparative Analysis of Imputation Methods for Single-Cell RNA Sequencing Data

Background: Single-cell RNA sequencing (scRNA-seq) enables gene expression profiling at cellular resolution but is inherently affected by sparsity caused by dropout events, where expressed genes are recorded as zeros due to technical limitations. These artifacts distort gene expression distributions and compromise downstream analyses. Numerous imputation methods have been proposed to recover latent transcriptional signals. These methods range from traditional statistical models to deep learning (DL)-based methods. However, their comparative performance remains unclear, as existing benchmarks evaluate only a limited subset of methods, datasets, and downstream analyses. Results: We present a comprehensive benchmark of 15 scRNA-seq imputation methods spanning 7 methodological categories, including traditional and DL-based methods. Methods are evaluated across 30 datasets from 10 experimental protocols on 6 downstream analyses. Results show that traditional methods, such as model-based, smoothing-based, and low-rank matrix-based methods, generally outperform DL-based methods, including diffusion-based, GAN-based, GNN-based, and autoencoder-based methods. In addition, strong performance in numerical gene expression recovery does not necessarily translate into improved biological interpretability in downstream analyses, including cell clustering, differential expression analysis, marker gene analysis, trajectory analysis, and cell type annotation. Furthermore, method performance varies substantially across datasets, protocols, and downstream analyses, with no single method consistently outperforming others. Conclusions: Our findings provide practical guidance for selecting imputation methods tailored to specific analytical objectives and underscore the importance of task-specific evaluation when assessing imputation performance in scRNA-seq data analysis.

q-bio.GN↗

arXiv2026

CDMT-EHR: A Continuous-Time Diffusion Framework for Generating Mixed-Type Time-Series Electronic Health Records

Electronic health records (EHRs) are invaluable for clinical research, yet privacy concerns severely restrict data sharing. Synthetic data generation offers a promising solution, but EHRs present unique challenges: they contain both numerical and categorical features that evolve over time. While diffusion models have demonstrated strong performance in EHR synthesis, existing approaches predominantly rely on discrete-time formulations, which suffer from finite-step approximation errors and coupled training-sampling step counts. We propose a continuous-time diffusion framework for generating mixed-type time-series EHRs with three contributions: (1) continuous-time diffusion with a bidirectional gated recurrent unit backbone for capturing temporal dependencies, (2) unified Gaussian diffusion via learnable continuous embeddings for categorical variables, enabling joint cross-feature modeling, and (3) a factorized learnable noise schedule that adapts to per-feature-per-timestep learning difficulties. Experiments on two large-scale intensive care unit datasets demonstrate that our method outperforms existing approaches in downstream task performance, distribution fidelity, and discriminability, while requiring only 50 sampling steps compared to 1,000 for baseline methods. Classifier-free guidance further enables effective conditional generation for class-imbalanced clinical scenarios.

cs.LG↗

arXiv2026

From Pen Strokes to Sleep States: Detecting Low-Recovery Days Using Sigma-Lognormal Handwriting Features

While handwriting has traditionally been studied for character recognition and disease classification, its potential to reflect day-to-day physiological fluctuations in healthy individuals remains unexplored. This study examines whether daily variations in sleep-related recovery states can be inferred from online handwriting dynamics. % We propose a personalized binary classification framework that detects low-recovery days using features derived from the Sigma-Lognormal model, which captures the neuromotor generation process of pen strokes. In a 28-day in-the-wild study involving 13 university students, handwriting was recorded three times daily, and nocturnal cardiac indicators were measured using a wearable ring. For each participant, the lowest (or highest) quartile of four sleep-related metrics -- HRV, lowest heart rate, average heart rate, and total sleep duration -- defined the positive class. Leave-One-Day-Out cross-validation showed that PR-AUC significantly exceeded the baseline (0.25) for all four variables after FDR correction, with the strongest performance observed for cardiac-related variables. Importantly, classification performance did not differ significantly across task types or recording timings, indicating that recovery-related signals are embedded in general movement dynamics. These results demonstrate that subtle within-person autonomic recovery fluctuations can be detected from everyday handwriting, opening a new direction for non-invasive, device-independent health monitoring.

cs.HC↗

arXiv2026

Prediction of Grade, Gender, and Academic Performance of Children and Teenagers from Handwriting Using the Sigma-Lognormal Model

Digital handwriting acquisition enables the capture of detailed temporal and kinematic signals reflecting the motor processes underlying writing behavior. While handwriting analysis has been extensively explored in clinical or adult populations, its potential for studying developmental and educational characteristics in children remains less investigated. In this work, we examine whether handwriting dynamics encode information related to student characteristics using a large-scale online dataset collected from Japanese students from elementary school to junior high school. We systematically compare three families of handwriting-derived features: basic statistical descriptors of kinematic signals, entropy-based measures of variability, and parameters obtained from the sigma-lognormal model. Although the dataset contains dense stroke-level recordings, features are aggregated at the student level to enable a controlled comparison between representations. These features are evaluated across three prediction tasks: grade prediction, gender classification, and academic performance classification, using Linear or Logistic Regression and Random Forest models under consistent experimental settings. The results show that handwriting dynamics contain measurable signals related to developmental stage and individual differences, especially for the grade prediction task. These findings highlight the potential of kinematic handwriting analysis and confirm that through their development, children's handwriting evolves toward a lognormal motor organization.

cs.HC↗

arXiv2026

Towards Non-Latin Text and Layout Personalization for Enhanced Readability

Reading has always been an integral part of both professional and personal life. Character and layout recognition and understanding by computers are well-explored areas. Nevertheless, how characters and layout are read and perceived by humans remains relatively underexplored. This work contributes to the field of human-document interaction (HDI) by investigating the effects of character and layout personalization on readability. The paper presents an empirical study on how parts-of-speech (POS)-based character and layout modifications can lead to overall improvements in both reading comprehension and memorization for two non-segmented, non-Latin writing systems: Khmer and Japanese. The experimental results from 43 participants suggest that, by bolding POS-derived content words, Khmer readers perform better on both reading comprehension and memorisation tasks, with a significant effect (p-values of 0.03 and 0.04, respectively). A similar overall tendency is also observed in a pilot study among Japanese readers (10 participants) using syntactic color-coding. In addition, the analyses of reading time, answering time, and perceived difficulty reveal that the proposed text styling technique does not increase any perceived difficulty, cognitive load, or reading effort for the Khmer readers. However, the Japanese readers experienced a decrease in reading speed. This study and its findings represent a significant step towards enabling dynamic, script-dependent personalization of character and layout to optimize human readability.

cs.HC↗

arXiv2026

Towards Universal Khmer Text Recognition

Khmer is a low-resource language characterized by a complex script, presenting significant challenges for optical character recognition (OCR). While document printed text recognition has advanced because of available datasets, performance on other modalities, such as handwritten and scene text, remains limited by data scarcity. Training modality-specific models for each modality does not allow cross-modality transfer learning, from which modalities with limited data could otherwise benefit. Moreover, deploying many modality-specific models results in significant memory overhead and requires error-prone routing each input image to the appropriate model. On the other hand, simply training on a combined dataset with a non-uniform data distribution across different modalities often leads to degraded performance on underrepresented modalities. To address these, we propose a universal Khmer text recognition (UKTR) framework capable of handling diverse text modalities. Central to our method is a novel modality-aware adaptive feature selection (MAFS) technique designed to adapt visual features according to a particular input image modality and enhance recognition robustness across modalities. Extensive experiments demonstrate that our model achieves state-of-the-art (SoTA) performance. Furthermore, we introduce the first comprehensive benchmark for universal Khmer text recognition, which we release to the community to facilitate future research. Our datasets and models can be accessible via this gated repository\footnote{in review}.

cs.CV↗

arXiv2026

Towards Khmer Scene Document Layout Detection

While document layout analysis for Latin scripts has advanced significantly, driven by the advent of large multimodal models (LMMs), progress for the Khmer language remains constrained because of the scarcity of annotated training data. This gap is particularly acute for scene documents, where perspective distortions and complex backgrounds challenge traditional methods. Given the structural complexities of Khmer script, such as diacritics and multi-layer character stacking, existing Latin-based layout analysis models fail to accurately delineate semantic layout units, particularly for dense text regions (e.g., list items). In this paper, we present the first comprehensive study on Khmer scene document layout detection. We contribute a novel framework comprising three key elements: (1) a robust training and benchmarking dataset specifically for Khmer scene layouts; (2) an open-source document augmentation tool capable of synthesizing realistic scene documents to scale training data; and (3) layout detection baselines utilizing YOLO-based architectures with oriented bounding boxes (OBB) to handle geometric distortions. To foster further research in the Khmer document analysis and recognition (DAR) community, we release our models, code, and datasets in this gated repository (in review).

cs.CV↗

arXiv2026

CBEN -- A Multimodal Machine Learning Dataset for Cloud Robust Remote Sensing Image Understanding

Clouds are a common phenomenon that distorts optical satellite imagery, which poses a challenge for remote sensing. However, in the literature cloudless analysis is often performed where cloudy images are excluded from machine learning datasets and methods. Such an approach cannot be applied to time sensitive applications, e.g., during natural disasters. A possible solution is to apply cloud removal as a preprocessing step to ensure that cloudfree solutions are not failing under such conditions. But cloud removal methods are still actively researched and suffer from drawbacks, such as generated visual artifacts. Therefore, it is desirable to develop cloud robust methods that are less affected by cloudy weather. Cloud robust methods can be achieved by combining optical data with radar, a modality unaffected by clouds. While many datasets for machine learning combine optical and radar data, most researchers exclude cloudy images. We identify this exclusion from machine learning training and evaluation as a limitation that reduces applicability to cloudy scenarios. To investigate this, we assembled a dataset, named CloudyBigEarthNet (CBEN), of paired optical and radar images with cloud occlusion for training and evaluation. Using average precision (AP) as the evaluation metric, we show that state-of-the-art methods trained on combined clear-sky optical and radar imagery suffer performance drops of 23-33 percentage points when evaluated on cloudy images. We then adapt these methods to cloudy optical data during training, achieving relative improvement of 17.2-28.7 percentage points on cloudy test cases compared with the original approaches. Code and dataset are publicly available at: https://github.com/mstricker13/CBEN

cs.CV↗

arXiv2025

Exploring the Alignment of Perceived and Measured Sleep Quality with Working Memory using Consumer Wearables

Wearable devices offer detailed sleep-tracking data. However, whether this information enhances our understanding of sleep or simply quantifies already-known patterns remains unclear. This work explores the relationship between subjective sleep self-assessments and sensor data from an Oura ring over 4--8 weeks in-the-wild. 29 participants rated their sleep quality daily compared to the previous night and completed a working memory task. Our findings reveal that differences in REM sleep, nocturnal heart rate, N-Back scores, and bedtimes highly predict sleep self-assessment in significance and effect size. For N-Back performance, REM sleep duration, prior night's REM sleep, and sleep self-assessment are the strongest predictors. We demonstrate that self-report sensitivity towards sleep markers differs among participants. We identify three groups, highlighting that sleep trackers provide more information gain for some users than others. Additionally, we make all experiment data publicly available.

cs.HC↗

arXiv2025

AI Solutionism and Digital Self-Tracking with Wearables

Self-tracking technologies and wearables automate the process of data collection and insight generation with the support of artificial intelligence systems, with many emerging studies exploring ways to evolve these features further through large-language models (LLMs). This is done with the intent to reduce capture burden and the cognitive stress of health-based decision making, but studies neglect to consider how automation has stymied the agency and independent reflection of users of self-tracking interventions. In this position paper, we explore the consequences of automation in self-tracking by relating it to our experiences with investigating the Oura Ring, a sleep wearable, and navigate potential remedies.

cs.HC↗

arXiv2025

Had enough of experts? Quantitative knowledge retrieval from large language models

Large language models (LLMs) have been extensively studied for their abilities to generate convincing natural language sequences, however their utility for quantitative information retrieval is less well understood. Here we explore the feasibility of LLMs as a mechanism for quantitative knowledge retrieval to aid two data analysis tasks: elicitation of prior distributions for Bayesian models and imputation of missing data. We introduce a framework that leverages LLMs to enhance Bayesian workflows by eliciting expert-like prior knowledge and imputing missing data. Tested on diverse datasets, this approach can improve predictive accuracy and reduce data requirements, offering significant potential in healthcare, environmental science and engineering applications. We discuss the implications and challenges of treating LLMs as 'experts'.

cs.IR↗

arXiv2025

A Review of Cognitive Readiness, Wearable Devices, and Prospects

In Human-Computer Interaction (HCI) and Ubiquitous Computing, the objective of optimizing device interactions and personalizing user experiences has placed a new emphasis on accurately evaluating cognitive readiness using wearable devices. Interpreting cognitive readiness in real-world scenarios is complex due to the plethora of potential physiological measures, individual variability, and the limitations of wearable devices. In this review, we present a systematic overview of key physiological measures that can be used for an in-depth assessment of cognitive readiness. These measures can serve as proxies for detailed assessments of cognitive readiness. This review serves as a tool for assessing cognitive readiness for diverse applications, with special focus on in-the-wild research settings. In addition, due to the complexity of measurements and devices, we propose the development of robust catalog for cognitive readiness measurements.

cs.HC↗

arXiv2024

Identifying Periods of Cyclical Stress in University Students Using Wearables In-the-Wild

University students encounter various forms of stress during their academic journey, including cyclical stress associated with final exams. Supporting their well-being means helping them manage their stress levels. In this study, we used a wearable health-tracking ring on a cohort of 103 Japanese university students for up to 28 months in the wild. The study aimed to investigate whether group-wide biomarkers of stress can be identified in a sample having similar daily schedules and whether these occurrences can be pinpointed to specific periods of the academic year. We found population-wide increased stress markers during exams, New Year's, and job hunting season, a Japanese job market peculiarity. Our results highlight the available potential of unobtrusive, in-situ detection of the current mental state of university student populations using off-the-shelf wearables from noisy data, with significant implications for the well-being of the users. Our approach and method of analysis allows for monitoring the student body's stress level without singling out individuals and therefore represents a privacy-preserving method. This way, new and sudden stress increases can be recognized, which can help identify the stressor and inform the design and introduction of counter measures.

cs.CY↗

arXiv2021

Confidence-Aware Learning Assistant

Not only correctness but also self-confidence play an important role in improving the quality of knowledge. Undesirable situations such as confident incorrect and unconfident correct knowledge prevent learners from revising their knowledge because it is not always easy for them to perceive the situations. To solve this problem, we propose a system that estimates self-confidence while solving multiple-choice questions by eye tracking and gives feedback about which question should be reviewed carefully. We report the results of three studies measuring its effectiveness. (1) On a well-controlled dataset with 10 participants, our approach detected confidence and unconfidence with 81% and 79% average precision. (2) With the help of 20 participants, we observed that correct answer rates of questions were increased by 14% and 17% by giving feedback about correct answers without confidence and incorrect answers with confidence, respectively. (3) We conducted a large-scale data recording in a private school (72 high school students solved 14,302 questions) to investigate effective features and the number of required training samples.

cs.HC↗

arXiv2020

Self-supervised Deep Learning for Reading Activity Classification

Reading analysis can give important information about a user's confidence and habits and can be used to construct feedback to improve a user's reading behavior. A lack of labeled data inhibits the effective application of fully-supervised Deep Learning (DL) for automatic reading analysis. In this paper, we propose a self-supervised DL method for reading analysis and evaluate it on two classification tasks. We first evaluate the proposed self-supervised DL method on a four-class classification task on reading detection using electrooculography (EOG) glasses datasets, followed by an evaluation of a two-class classification task of confidence estimation on answers of multiple-choice questions (MCQs) using eye-tracking datasets. Fully-supervised DL and support vector machines (SVMs) are used to compare the performance of the proposed self-supervised DL method. The results show that the proposed self-supervised DL method is superior to the fully-supervised DL and SVM for both tasks, especially when training data is scarce. This result indicates that the proposed self-supervised DL method is the superior choice for reading analysis tasks. The results of this study are important for informing the design and implementation of automatic reading analysis platforms.

cs.HC↗

arXiv2020

Distortion-Adaptive Grape Bunch Counting for Omnidirectional Images

This paper proposes the first object counting method for omnidirectional images. Because conventional object counting methods cannot handle the distortion of omnidirectional images, we propose to process them using stereographic projection, which enables conventional methods to obtain a good approximation of the density function. However, the images obtained by stereographic projection are still distorted. Hence, to manage this distortion, we propose two methods. One is a new data augmentation method designed for the stereographic projection of omnidirectional images. The other is a distortion-adaptive Gaussian kernel that generates a density map ground truth while taking into account the distortion of stereographic projection. Using the counting of grape bunches as a case study, we constructed an original grape-bunch image dataset consisting of omnidirectional images and conducted experiments to evaluate the proposed method. The results show that the proposed method performs better than a direct application of the conventional method, improving mean absolute error by 14.7% and mean squared error by 10.5%.

cs.CV↗