The field of agricultural research is witnessing a significant shift towards the adoption of artificial intelligence (AI) and machine learning (ML) technologies. Recent developments have focused on improving crop disease detection, livestock health management, and aquatic species monitoring through the use of deep learning techniques and vision-language models. Notably, large language models tailored for specific domains such as aquaculture and agriculture are being developed to support farmers, researchers, and industry practitioners. Furthermore, benchmarks for evaluating the performance of vision-language models in agricultural tasks have been proposed, highlighting the need for more accurate and scalable solutions.

Some noteworthy papers in this area include: AQUA, a large language model for aquaculture, which has the potential to drive innovations in aquaculture research and decision-making tools. AgroBench, a benchmark for evaluating vision-language models in agriculture, which reveals that current models have room for improvement in fine-grained identification tasks.