A Comparative Analysis between Visual Language and Large Language Models to Monitor Contaminated Water in the face of Climate Change

In recent years, Artificial Intelligence (AI) has advanced at an unprecedented rate, showing promising potential in transforming industries and individual lifestyles. Our research focused on a comparative analysis between three approaches to report generation based on prompts using the Visual Language Model (VLM) Llama-3.2-11b-vision-preview and the Large Language Model (LLM) Llama-3.1-8b-instant. The VLM processed images, and the LLM handled text prompts. Both models were later combined for an innovative report-generation approach. The study indicated that the combined approach proved superior due to the integration of both visual and textual features. However, it is important to note that the performance of AI models heavily depends on factors such as data quality, computational resources, and prompts. By addressing these factors, this research also facilitates advanced monitoring of water quality, supporting sustainability efforts, and enhancing climate resilience.