Browse Articles
Discover research articles across all indexed journals
EEG-based automated evaluation of automotive sound quality using ensemble deep learning
Abstract The evaluation of automotive sound quality is of considerable significance for improving driving comfort. However, existing methodologies suffer from notable limitations, including inconsistencies in subjective evaluations and weak correlations between objective metrics and auditory perception. In response to these challenges, an automated evaluation method incorporating electroencephalogram (EEG) signals and ensemble deep learning is proposed herein. Initially, EEG data is acquired from 30 subjects during exposure to 16 automobile sounds with sporty quality. Subsequently, the LSTMS-B model is incorporating Swish activation into LSTM to mitigate gradient vanishing and enhancing Bagging through optimized majority voting, achieving 90.8% accuracy with superior performance over conventional LSTM variants; Furthermore, an innovative ResNet-based regression model is developed to establish the automobile sound-EEG feature mapping, enabling the LSTMS-B model to achieve 89.75% average F1 score in sound quality classification using brain auditory representations while reducing reliance on conventional EEG paradigms. This study develops a novel sound quality evaluation paradigm through deep-ensemble learning integration, where the proposed cross-modal feature mapping method provides a transferable AI framework for interpreting human auditory perception mechanisms.
Automated quantification of stereotypical motor movements in autism using persistent homology
BCG vaccination at birth shapes the TCR usage and functional gene expression profile of MR1T cells at 9 weeks of age
Exploring the role of a social robot in building a sense of connection
Abstract An important but under-explored use case for socially assistive robotics is in community formation. In this work, we conduct a novel investigation into the ability of a social robot to encourage human-human relationships among users. We conduct a between-subjects Wizard-of-Oz experiment in which we focus on the robot’s role in interaction as a determinant in the sense of connection that participants feel as a result of interacting with the robot. N=86 college-age participants were recruited to watch a video with a robot, where the robot’s group membership and behavior were adjusted to reflect different roles. We vary the robot’s group membership along three levels: the same group as the participants, engaging in the same activity as the participants, or simply being in the same room and acknowledging the participants, with a control condition of no robot present. Before, during, and after the video, the robot would speak to encourage interaction between participants and the robot when applicable. After the video was concluded, participant sense of connection was measured through a series of Likert surveys. We find that inter-user conversation increased when the robot actively participated in the video viewing and asked questions of the users compared to when the robot did not. We also find some exploratory evidence that these effects might extend to user attitudes about not only each other but also the laboratory running the experiment.
Veno-arterial ECMO as a bridge to delayed repair in post-infarction ventricular septal rupture with cardiogenic shock
Modeling of the electrostatic field generated by horizontal current sources in a sea-ice-covered four-layer medium
Hydrogenation of nitriles to primary amines by a phosphine-free cobalt(II) complex in the absence of H2 gas
Exploratory analysis of genomic prediction profiles in Arequipa fighting cattle using commercial SNP panels
A novel hydrophobic deep eutectic solvent-based paired-drop microextraction method for the trace determination of metribuzin in urine samples
Abstract A hydrophobic deep eutectic solvent (DES) was molecularly tailored from tributylphosphine oxide and octanoic acid via hydrogen-bonding and employed as an efficient extractant in a novel ultrasound-assisted paired-drop microextraction (UAE-PDME) method for metribuzin determination in urine. The DES structure was confirmed by FTIR analysis. The innovative paired-drop design addresses a key practical challenge: while ultrasound efficiently disperses the extractant, quantitatively retrieving a minute volume (e.g., 2 µL) of dispersed solvent from an aqueous sample is nearly impossible. This method solves this by using a second, hanging DES droplet that coalesces with the dispersed microdroplets during gentle stirring, enabling efficient, complete, and centrifugation-free collection of the enriched phase. After optimization of key parameters, the method was rigorously validated using spiked urine samples. It demonstrated excellent performance: a wide linear range of 5–500 µg/L, low limits of detection and quantification of 0.20 and 0.67 µg/L, respectively, high precision (RSD < 3.6%), quantitative recoveries (99.5–103.1%) with negligible matrix effect (− 4.2%) and high solvent phase recovery (98.5%). The high extraction efficiency is attributed to the DES’s rational molecular design. The DES-UAE-PDME method represents a green, rational, and effective sample preparation strategy for trace pesticide monitoring using conventional HPLC–UV instrumentation.
Effect of biotic and abiotic degradation on the surface and bulk of rubber compounds
Linking sensory processing patterns and somatosensory thresholds to behavioural and emotional symptoms in adolescents with ADHD
Synthesis, spectroscopic characterization and computational evaluation of 2-alkylthio imidazol-4-one derivatives as potential CDK5 inhibitors
Recognition of everyday activities using experiment data from wearable sensors: a deep learning-based framework
Abstract Tracking everyday activities is vital for detecting changes in older adults’ health, allowing timely support to promote well-being. Wearable sensors and deep learning provide continuous monitoring, making them a supportive tool in detecting such changes. However, a more refined method is needed to recognise precise activities with a minimal set of sensors. This study aimed to develop a method to recognise everyday activities among older adults by utilising wearable sensors and a deep learning model. This is a small-scale home lab experiment to develop a method to recognise 14 everyday activities. We compared five models that recognised everyday activities with different sensor signal counts and accuracy. Our results showed that sensor placement is important. Based on the results, we proposed a two-sensor method (pelvis and right hand) to collect and correctly recognise everyday activities among older adults. This model, which utilises two sensors, classified 12 activities with an accuracy of 89.3%. Another model recognised all 14 activities with a lower accuracy of 88.2% using five sensors. We also explored a one-sensor approach, which showed low recognition performance and struggled to distinguish activity variability. The two-sensor-based system will allow for large-scale data collection on everyday activities of older adults.
Development and application of a poly (sodium p-styrenesulfonate-co-acrylic acid) / Ficus leaves–derived activated carbon composite for the efficient removal of 4-nitrophenol from aqueous solutions
Abstract This research details the synthesis and application of a novel poly (sodium p-styrenesulfonate-co-acrylic acid)/Ficus leaf–derived activated carbon composite (P(NaSS-co-AA)/AC) for the Efficient elimination of 4-Nitrophenol from contaminated Aqueous Solutions. The copolymer P(NaSS-co-AA) was fabricated via free-radical copolymerization of sodium p-styrenesulfonate (NaSS) and acrylic acid (AA) and characterized by GPC, ¹H NMR, and ¹³C NMR. The composite was further analyzed utilizing SEM, FTIR, BET, and TGA techniques. Batch adsorption tests experiments were performed to optimize operational parameters achieving optimal conditions at pH 7, an adsorbent dosage of 1.5 g/L, and a contact time of 120 min, with a removal efficiency of 95 ± 1.18%. Adsorption followed the Langmuir isotherm and pseudo-second-order kinetics, suggesting monolayer chemisorption on a homogeneous surface. The composite demonstrated high selectivity and robust performance in real wastewater matrices and multicomponent ionic environments, maintaining a removal efficiency of over 90% despite the presence of competing species. The maximum adsorption capacity was 65.0 ± 1.25 mg/g, exceeding conventional activated carbons. Statistical evaluation of models showed excellent fits (Langmuir R² = 0.994, RMSE = 0.85 ± 0.08; kinetics pseudo-second-order R² = 0.999, RMSE = 0.47 ± 0.04). Post-adsorption FTIR and XPS analyses confirmed 4-NP uptake and supported the proposed multi-mode adsorption mechanism (π–π stacking, electrostatic attraction, and hydrogen bonding). Preliminary column breakthrough experiments and leaching/environmental safety screening further validate the practical potential of this composite. These results showed that the developed composite is a sustainable and cost-effective adsorbent with strong potential for wastewater treatment applications targeting nitrophenol contaminants.