A Method Based on Knowledge Distillation for Fish School Stress State Recognition in Intensive Aquaculture
Siyuan Mei1,2, Yingyi Chen1,2,*, Hanxiang Qin1,2, Huihui Yu3, Daoliang Li1,2, Boyang Sun1,2, Ling Yang1,2, Yeqi Liu1,2
CMES-Computer Modeling in Engineering & Sciences, Vol.131, No.3, pp. 1315-1335, 2022, DOI:10.32604/cmes.2022.019378
- 19 April 2022
Abstract Fish behavior analysis for recognizing stress is very important for fish welfare and production management in aquaculture. Recent advances have been made in fish behavior analysis based on deep learning. However, most existing methods with top performance rely on considerable memory and computational resources, which is impractical in the real-world scenario. In order to overcome the limitations of these methods, a new method based on knowledge distillation is proposed to identify the stress states of fish schools. The knowledge distillation architecture transfers additional inter-class information via a mixed relative loss function, and it forces a… More >