Metaheuristics Optimization with Deep Learning Enabled Automated Image Captioning System

Image captioning is a popular topic in the domains of computer vision and natural language processing (NLP). Recent advancements deep learning (DL) models have enabled improvement overall performance image approach. This study develops metaheuristic optimization with learning-enabled automated technique (MODLE-AICT). The proposed MODLE-AICT model focuses on generation effective captions to input images by using two processes involving encoding unit decoding unit. Initially, at part, salp swarm algorithm (SSA), HybridNet model, utilized generate effectual representation fixed-length vectors, showing novelty work. Moreover, part includes bidirectional gated recurrent (BiGRU) used descriptive sentences. inclusion an SSA-based hyperparameter optimizer helps attaining performance. For inspecting enhanced series simulations were carried out, results are examined under several aspects. experimental values suggested betterment over recent approaches.