Search arXiv⌕ Search

arXiv subjects

Riku Klen

Publications and source records attributed to Riku Klen.

6 recordsLinked to original sources
arXiv2026

Comparative Analysis of State-of-the-Art Foundation Models for Sleep Analysis Under Channel Reduction

Automatic sleep staging from polysomnography (PSG) is a well-studied task, but PSG itself is expensive, clinic-based, and burdensome to manually score, which limits its use for long-term or at-home monitoring. Most existing sleep-staging foundation models are evaluated using the full PSG montage. We instead ask how much of that montage is actually necessary. We evaluate six sleep staging models on the Multi-Ethnic Study of Atherosclerosis (MESA) PSG dataset across three signal conditions: electroencephalography (EEG), electrocardiography (ECG), and their combination (EEG+ECG). This is motivated by edge-cloud deployment, where EEG requires a clinic-grade scalp electrode, whereas ECG is already captured by consumer wearables. We test state-of-the-art foundation models such as SleepFM with an encoder trained from scratch on MESA, alongside BIOT, MOMENT, LaBraM, a base-scale Vision Transformer (ViT-B) reimplementation of SensorLM trained from scratch, and YASA, spanning EEG-pretrained, general-time-series, from-scratch, and classical non-learned approaches. No model architecture is modified from its original form; SensorLM's encoder is reimplemented only in PyTorch. For EEG-only staging, BIOT achieves the best result with a macro~F1 of 0.7237, followed by LaBraM (0.6835) and SleepFM from scratch (0.6582). Across the five models capable of ECG-only staging, switching from EEG to ECG costs between 0.2798 (MOMENT) and 0.4151 (BIOT) macro~F1, averaging 0.3531, while cutting the raw channel data rate to a third. Adding ECG to EEG provides no gain for most models. These results show that EEG carries most of the sleep-staging signal, quantify the consistent accuracy cost of the wearable-compatible alternative, and demonstrate that sleep-relevant pretraining transfers well to MESA.

eess.SP↗
arXiv2025

Emotion Recognition with Minimal Wearable Sensing: Multi-domain Feature, Hybrid Feature Selection, and Personalized vs. Generalized Ensemble Model Analysis

Negative emotions are linked to the onset of neurodegenerative diseases and dementia, yet they are often difficult to detect through observation. Physiological signals from wearable devices offer a promising noninvasive method for continuous emotion monitoring. In this study, we propose a lightweight, resource-efficient machine learning approach for binary emotion classification, distinguishing between negative (sadness, disgust, anger) and positive (amusement, tenderness, gratitude) affective states using only electrocardiography (ECG) signals. The method is designed for deployment in resource-constrained systems, such as Internet of Things (IoT) devices, by reducing battery consumption and cloud data transmission through the avoidance of computationally expensive multimodal inputs. We utilized ECG data from 218 CSV files extracted from four studies in the Psychophysiology of Positive and Negative Emotions (POPANE) dataset, which comprises recordings from 1,157 healthy participants across seven studies. Each file represents a unique subject emotion, and the ECG signals, recorded at 1000 Hz, were segmented into 10-second epochs to reflect real-world usage. Our approach integrates multidomain feature extraction, selective feature fusion, and a voting classifier. We evaluated it using a participant-exclusive generalized model and a participant-inclusive personalized model. The personalized model achieved the best performance, with an average accuracy of 95.59%, outperforming the generalized model, which reached 69.92% accuracy. Comparisons with other studies on the POPANE and similar datasets show that our approach consistently outperforms existing methods. This work highlights the effectiveness of personalized models in emotion recognition and their suitability for wearable applications that require accurate, low-power, and real-time emotion tracking.

cs.HC↗
arXiv2025

Improved Brain Tumor Detection in MRI: Fuzzy Sigmoid Convolution in Deep Learning

Early detection and accurate diagnosis are essential to improving patient outcomes. The use of convolutional neural networks (CNNs) for tumor detection has shown promise, but existing models often suffer from overparameterization, which limits their performance gains. In this study, fuzzy sigmoid convolution (FSC) is introduced along with two additional modules: top-of-the-funnel and middle-of-the-funnel. The proposed methodology significantly reduces the number of trainable parameters without compromising classification accuracy. A novel convolutional operator is central to this approach, effectively dilating the receptive field while preserving input data integrity. This enables efficient feature map reduction and enhances the model's tumor detection capability. In the FSC-based model, fuzzy sigmoid activation functions are incorporated within convolutional layers to improve feature extraction and classification. The inclusion of fuzzy logic into the architecture improves its adaptability and robustness. Extensive experiments on three benchmark datasets demonstrate the superior performance and efficiency of the proposed model. The FSC-based architecture achieved classification accuracies of 99.17%, 99.75%, and 99.89% on three different datasets. The model employs 100 times fewer parameters than large-scale transfer learning architectures, highlighting its computational efficiency and suitability for detecting brain tumors early. This research offers lightweight, high-performance deep-learning models for medical imaging applications.

eess.IV↗
arXiv2007

Local Convexity Properties of j-metric Balls

This paper deals with local convexity properties of the j-metric. We consider convexity and starlikeness of the j-metric balls in convex, starlike and general subdomains of R^n.

math.MG↗