Improving facial expression recognition for autism with IDenseNet‐RCAformer under occlusions DOI

S. Selvi,

M. Parvathy

International Journal of Developmental Neuroscience, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 27, 2024

The term 'autism spectrum disorder' describes a neurodevelopmental illness typified by verbal and nonverbal interaction impairments, repetitive behaviour patterns poor social interaction. Understanding mental states from FEs is crucial for interpersonal But when there are occlusions like glasses, facial hair or self-occlusion, it becomes harder to identify expressions accurately. This research tackles the issue of identifying parts face occluded suggests an innovative technique tackle this difficulty. Creating strong framework expression recognition (FER) that better handles increases accuracy goal research. Therefore, we propose novel Improved DenseNet-based Residual Cross-Attention Transformer (IDenseNet-RCAformer) system partial occlusion FER problem in autism patients. framework's efficacy assessed using four datasets expressions, some preprocessing procedures conducted increase efficiency. After that, recognizing simple argmax function applied get forecasted landmark position heatmap. Then feature extraction phase, local global representation captured preprocessed images adopting Inception-ResNet-V2 approach, Transformer, respectively. Moreover, both features fused employing FusionNet method, thereby enhancing system's training speed precision. extracted, improved DenseNet mechanism efficiently recognize variety partially A number performance metrics determined analysed demonstrate proposed approach's effectiveness, where IDenseNet-RCAformer performs best with 98.95%. According experimental findings, significantly outperforms prior frameworks terms outcomes.

Язык: Английский

Enhanced handwriting recognition through hybrid UNet-based architecture with global classical features DOI
Xiaofei Liu

Journal of Ambient Intelligence and Humanized Computing, Год журнала: 2025, Номер unknown

Опубликована: Янв. 18, 2025

Язык: Английский

Процитировано

0
Development of a framework using deep learning for the identification and classification of engagement levels in distance learning students DOI Creative Commons
Fernando Rodrigues Trindade Ferreira,

Loena Marins do Couto,

Guilherme de Melo Baptista Domingues

и другие.

Social Network Analysis and Mining, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 11, 2025

Язык: Английский

Процитировано

0
Optimized SVR with nature-inspired algorithms for environmental modelling of mycotoxins in food virtual-water samples DOI Creative Commons
A. G. Usman, Sagiru Mati, Hanita Daud

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Май 13, 2025

The accurate determination of mycotoxins in food samples is crucial to guarantee safety and minimize their toxic effects on human animal health. This study proposed the use a support vector regression (SVR) predictive model improved by two metaheuristic algorithms used for optimization namely, Harris Hawks Optimization (HHO) Particle Swarm (PSO) predict chromatographic retention time various mycotoxin groups. dataset was collected from secondary sources train validate SVR-HHO SVR-PSO models. performance models assessed via mean square error, correlation coefficient, Nash-Sutcliffe efficiency. outperformed existing methods 4-7% both learning (training testing) phases respectively. By using optimization, parameter adjustment became more effective, avoiding trapping local minima improving generalization. These results demonstrate how machine metaheuristics may be combined accurately forecast levels, providing useful tool regulatory compliance monitoring. framework perfect commercial quality assurance, testing, extensive programs because it provides exceptional accuracy resilience predicting times. In contrast conventional models, effectively manages intricate nonlinear interactions, guaranteeing identification while lowering hazards

Язык: Английский

Процитировано

0
Optimizing Wind Power Forecasting with RNN-LSTM Models through Grid Search Cross-validation DOI

Ahmed Mohamed Reda Abdelkader,

Hanaa ZainEldin, Mahmoud M. Saafan

и другие.

Sustainable Computing Informatics and Systems, Год журнала: 2024, Номер unknown, С. 101054 - 101054

Опубликована: Ноя. 1, 2024

Язык: Английский

Процитировано

2
Optimization of Hydropower Unit Startup Process Based on the Improved Multi-Objective Particle Swarm Optimization Algorithm DOI Creative Commons
Qingquan Zhang,

Zifeng Xie,

Mingming Lu

и другие.

Energies, Год журнала: 2024, Номер 17(17), С. 4473 - 4473

Опубликована: Сен. 6, 2024

In order to improve the dynamic performance during startup process of hydropower units, while considering efficient and stable speed increase effective suppression water pressure fluctuations mechanical vibrations, optimization algorithms must be used select optimal parameters for system. However, in current research, various multi-objective still have limitations terms target space coverage diversity maintenance parameter hydraulic turbines. To explore verify turbines, multiple strategies are proposed this study. Under condition constructing a fine-tuned nonlinear model control system, paper focuses on three key indicators: absolute integral deviation, snail shell fluctuation, relative value maximum axial thrust. Through comparative analysis particle swarm algorithm (MOPSO), variant (VMOPSO), sine cosine (MOSCA), biogeography (MOBBO), gravity search (MOGAS), improved (IMOPSO), obtained compared analyzed strategy, most suitable actual working conditions selected through comprehensive weighting method. The results show that, local solution problem caused by other algorithms, method significantly reduces vibrations ensuring improvement, achieving better performance. significant guiding significance smooth operation safety provide strong support making operational decisions.

Язык: Английский

Процитировано

1
Optimized YOLOV8: An efficient underwater litter detection using deep learning DOI Creative Commons

Faiza Rehman,

Mariam Rehman,

Maria Anjum

и другие.

Ain Shams Engineering Journal, Год журнала: 2024, Номер 16(1), С. 103227 - 103227

Опубликована: Дек. 26, 2024

Язык: Английский

Процитировано

1
Improving facial expression recognition for autism with IDenseNet‐RCAformer under occlusions DOI

S. Selvi,

M. Parvathy

International Journal of Developmental Neuroscience, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 27, 2024

The term 'autism spectrum disorder' describes a neurodevelopmental illness typified by verbal and nonverbal interaction impairments, repetitive behaviour patterns poor social interaction. Understanding mental states from FEs is crucial for interpersonal But when there are occlusions like glasses, facial hair or self-occlusion, it becomes harder to identify expressions accurately. This research tackles the issue of identifying parts face occluded suggests an innovative technique tackle this difficulty. Creating strong framework expression recognition (FER) that better handles increases accuracy goal research. Therefore, we propose novel Improved DenseNet-based Residual Cross-Attention Transformer (IDenseNet-RCAformer) system partial occlusion FER problem in autism patients. framework's efficacy assessed using four datasets expressions, some preprocessing procedures conducted increase efficiency. After that, recognizing simple argmax function applied get forecasted landmark position heatmap. Then feature extraction phase, local global representation captured preprocessed images adopting Inception-ResNet-V2 approach, Transformer, respectively. Moreover, both features fused employing FusionNet method, thereby enhancing system's training speed precision. extracted, improved DenseNet mechanism efficiently recognize variety partially A number performance metrics determined analysed demonstrate proposed approach's effectiveness, where IDenseNet-RCAformer performs best with 98.95%. According experimental findings, significantly outperforms prior frameworks terms outcomes.

Язык: Английский

Процитировано

0