On the evening of July 8, Associate Professor Feng Zaiwen from the School of Information at Huazhong Agricultural University delivered a presentation titled “Transformer Model and Its Application in Smart Agriculture” at the seminar.

Professor Feng provided a comprehensive overview of the Transformer model, tracing its origins, development, and applications. Initially designed for natural language processing tasks such as text translation and intelligent knowledge Q&A, the Transformer model has since expanded into image processing, autonomous driving, and other fields. He highlighted the common branches within the Transformer family, including the Encoder branch (e.g., BERT) and the Decoder branch (e.g., GPT). Professor Feng delved into the self-attention mechanism of the decoder in the model structure, explaining its significance and functionality. He concluded by introducing some of the applications of the Transformer model in agriculture, such as crop breeding models and environmental control models for flowers, fruits, and vegetables.

In the subsequent discussion, participants explored the potential applications of the Transformer model in agricultural economics, including its use in causal analysis and price prediction.