Abstract

This research presents a novel approach for feature extraction in image understanding, utilizing an AT-ConvLSTM-based Cultural Algorithm. The Proposed CA-AT-ConvLSTM leverages the power of deep learning through AT-ConvLSTM architecture while optimizing the feature extraction process using Cultural Algorithms. This synergistic approach enhances the efficiency and accuracy of image understanding tasks, making it suitable for a wide range of applications, from computer vision to pattern recognition. The experimental results demonstrate the superiority of the proposed technique over traditional methods, highlighting its potential in advancing the field of image analysis.