Retrieval-Based Inception V3-Net Algorithm and Invariant Data Classification using Enhanced Deep Belief Networks for Content-Based Image Retrieval
Year: 2025
Type: article
Abstract: In the present scenario, Content-Based Image Retrieval (CBIR) performs a constantly changing function that makes use gain knowledge from images. Moreover, it is also the dynamic sector of research and was recently rewarded due to the drastic increase in the performance of digital images. To retrieve images from the massive dataset, experts utilize Content Based Image Retrieval. This approach automatically indexes and retrieves images depending upon the contents of the image, and the developing techniques for mining images are based on the CBIR systems. Based on the visual characteristics of the input image, object pattern, texture, color, shape, layout, and position classifications are applied, and indexing is carried out. When issues arise during feature extraction, deep learning approaches help to resolve them. A method called RIV3-NET, which stands for Retrieval-Based Inception V3, was used to classify the features. Classifying image invariant data using Enhanced Deep Belief Networks (EDBN) is necessary to decrease noise and improve displacement with smoothness. The simulation outcomes demonstrate the improved picture retrieval and parametric analysis.
Source: THE SCIENTIFIC TEMPER
Authors: Shaik Abdulla P., Abdul Razak T.
Institution: Bharathidasan University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4409600063