Effective Image Annotation with Self Attention Capsule Network and Adaptive Residual CNN with Bionic Features

Loading...
Thumbnail Image

Date

item.page.authors

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Leveraging advanced deep learning architectures has recently emerged, showcas ing remarkable performance in annotation tasks. Despite the impressive strides made in image annotation techniques, the challenge lies in ensuring scalability and general ization across diverse datasets and real-world applications. The novel approaches for image annotation are proposed and evaluated. Stacking the BiLSTM approach com bined with a capsule network, achieving an impressive accuracy rate of 99%. This approach leverages self-attention mechanisms to effectively record the contextual infor mation moreover improve annotation accuracy. The second method introduces an adap tive optimized image annotation, which incorporates residual convolution and feature selection via the bionic technique. This model demonstrates robust performance with overall precision scores of 56%, 99.2%, and 90% across different evaluation metrics. By combining deep learning architectures with innovative optimization strategies, both approaches contribute significantly to the field of image annotation, offering promising avenues for further research and application in diverse domains. In future, the integra tion of these advanced image annotation methods such as medical imaging will pave the way for accurate image analysis in various domains. newlineSelf-Attention Capsule Network, Convolutional Neural Network, Bionic Features, Adaptive Red Colobus Monkey Bionic Model, Stacked Bidirectional Capsule Network, Local Binary Pattern. newline1. Vikas Palekar, Sathish Kumar L, quotAn Effective Image Annotation Using Self Attention Based Stacked Bidirectional Capsule Networkquot, Computer Standards and Interfaces, 2025, 103973, ISSN 0920-5489, https://doi.org/10.1016/j.csi.2025.103973. (SCIE IF- 4.1 Q1) newline newline2. Vikas Palekar, Sathish Kumar L, quotAdaptive optimized residual convolutional im age annotation model with bionic feature selection modelquot, Computer Standards and Interfaces, Volume 87, 2024, 103780, ISSN 0920-5489, https://doi.org/10.1016/j.csi.2023.103780. (SCI IF- 4.1 Q1) newline newline newlineIEEE.(Scopus) newline newline3. Palekar, V.R. and Kumar, S., 2020. Survey and Analysis on Machine Learning Approaches for Image Annotation. Journal of University of Shanghai for Science and Technology, 22(12), pp.1070-1081. DOI:10.51201/jusst12465 newline newline1. Vikas Palekar and Dr. Satish Kumar. L, A Critical Analysis of Learning Approaches for Image Annotation Based on Semantic Correlation Second International Conference on Security, Privacy and Data Analytics (ISPDA-2022), December 13-15, 2022 newlinehttps://doi.org/10.1007/978-981-99-3569-7_28 (Scopus) newline newline2. Palekar, V. and Kumar, S., 2021, December. Label dependency classifier us ing multi feature graph convolution networks for automatic image annotation. In 2021International Conference on Computational Performance Evaluation (ComPE) (pp. 619-624). newline doi: 10.1109/ComPE53109.2021.9752236 newline newline newline

Description

Keywords

Citation

item.page.endorsement

item.page.review

item.page.supplemented

item.page.referenced