Paper Title
A ROBUST CONTENT-BASED IMAGE RETRIEVAL SYSTEM INTEGRATING SIFT AND ATTENTION-BASED CONVOLUTIONAL NEURAL NETWORKSAbstract
Content-Based Image Retrieval (CBIR) has emerged as a critical solution for managing and retrieving large-scale image datasets by utilizing visual features such as color, texture, and shape. However, conventional CBIR systems suffer from limitations including the semantic gap, inefficient feature extraction, and high computational complexity. To address these challenges, this study proposes an advanced CBIR framework integrating preprocessing, feature extraction, and deep learning-based classification. Initially, Histogram Equalization (HE) is employed to enhance image contrast and improve visual quality. Subsequently, Scale-Invariant Feature Transform (SIFT) is used to extract robust and invariant features. These features are then processed using an Attention-Enhanced Convolutional Neural Network (CNN), which dynamically focuses on salient regions of images to improve retrieval accuracy. The proposed approach effectively combines traditional feature extraction with deep learning and attention mechanisms, leading to improved retrieval performance and robustness. The framework demonstrates enhanced accuracy, reduced noise sensitivity, and better adaptability across diverse image datasets. This work contributes to the development of efficient, scalable, and intelligent CBIR systems suitable for real-world applications such as medical imaging, multimedia databases, and digital libraries.
KEYWORDS : Content-Based Image Retrieval (CBIR), Histogram Equalization (HE), SIFT, Attention-Enhanced CNN, Deep Learning, Image Retrieval, Feature Extraction, Semantic Gap
