Comparative Analysis of RNN Strategies for Image Caption Generation: Evaluating Accuracy and Efficiency Across Feature Extraction Models

Researchers in the Artificial Intelligence department have long held a strong interest in image caption generation. This area proves invaluable across fields such as robotic vision and business, where training computers to accurately describe images or environments is essential. Over the years, achieving this task has posed a significant challenge within artificial intelligence.Focusing on various RNN strategies and analyzing their influence on phrase production, this study introduces multiple deep neural network models designed for generating image captions. Additionally, the study includes the creation of captions for sample photos and compares several feature extraction and encoder models to determine which achieves the highest accuracy in caption generation.