Method of Adaptive Knowledge Distillation from Multi-Teacher to Student Deep Learning Models
Journal of Edge Computing, Vol. 4, No. 2, pp. 159-178
Abstract
This paper proposes a method of adaptive knowledge distillation from multiple teacher models to a student model for efficient deployment on edge devices. The approach dynamically weights teacher contributions based on their expertise for different input samples.
Citation
Pavlo Radiuk, Oleksandr Chaban, Eduard Manziuk. "Method of Adaptive Knowledge Distillation from Multi-Teacher to Student Deep Learning Models". Journal of Edge Computing, Vol. 4, No. 2, pp. 159-178, 2025.