CapsWSO ‐Net Deep Learning Architecture for Intelligent Electronic Waste Detection in Recycling Environments Applications

Accurate classification of electronic waste (E‐waste) is important for automated sorting and material recovery, but variations in object condition, orientation, illumination, and background make visual identification difficult. Conventional attention‐based convolutional neural networks emphasize informative image regions but do not explicitly preserve part‐whole relationships, while capsule networks retain spatial hierarchies but include irrelevant and redundant features. This paper proposes CapsWSO‐Net, an attention‐enhanced capsule framework with optimization‐based feature selection. The model combines Channel‐Augmented Attention and Spatial‐Augmented Attention to refine channel and spatial information, followed by Axial Squeeze Attention to capture horizontal and vertical contextual relationships. White Shark Optimization with Opposition‐Based Learning is then used to select informative features from the attention‐enhanced capsule representation. CapsWSO‐Net is evaluated using three publicly available E‐waste image datasets achieved accuracy of 98.88% and an F 1‐score of 98.31%. The results demonstrate the effectiveness of combining attention refinement, capsule representation, and optimized feature selection for E‐waste classification.

Authors

Institutions

Publication Details

Journal
Transactions on Emerging Telecommunications Technologies
Published
2026-11-20
DOI
https://doi.org/10.1002/ett.70512
Primary Topic
Advanced Neural Network Applications
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

CapsWSO ‐Net Deep Learning Architecture for Intelligent Electronic Waste Detection in Recycling Environments Applications

G. Manikandan, D. Jeya Priya
Transactions on Emerging Telecommunications Technologies
Advanced Neural Network Applications
article

CapsWSO ‐Net Deep Learning Architecture for Intelligent Electronic Waste Detection in Recycling Environments Applications

G. Manikandan, D. Jeya Priya
article en

Abstract

Accurate classification of electronic waste (E‐waste) is important for automated sorting and material recovery, but variations in object condition, orientation, illumination, and background make visual identification difficult. Conventional attention‐based convolutional neural networks emphasize informative image regions but do not explicitly preserve part‐whole relationships, while capsule networks retain spatial hierarchies but include irrelevant and redundant features. This paper proposes CapsWSO‐Net, an attention‐enhanced capsule framework with optimization‐based feature selection. The model combines Channel‐Augmented Attention and Spatial‐Augmented Attention to refine channel and spatial information, followed by Axial Squeeze Attention to capture horizontal and vertical contextual relationships. White Shark Optimization with Opposition‐Based Learning is then used to select informative features from the attention‐enhanced capsule representation. CapsWSO‐Net is evaluated using three publicly available E‐waste image datasets achieved accuracy of 98.88% and an F 1‐score of 98.31%. The results demonstrate the effectiveness of combining attention refinement, capsule representation, and optimized feature selection for E‐waste classification.

Transactions on Emerging Telecommunications TechnologiesVol. 37(12)
Saint Joseph's College (US), St. Joseph's Institute of Technology (IN), St. Joseph’s College of Engineering (IN)
Advanced Neural Network Applications
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon