The book provides a timely coverage of the paradigm of knowledge distillationan efficient way of model compression. Knowledge distillation is positioned in a general setting of transfer learning, which effectively learns a lightweight student model from a large teacher model. The book covers a variety of training schemes, teacherstudent architectures, and distillation algorithms. The book covers a wealth of topics including recent developments in vision and language learning, relational architectures, multi-task learning, and representative applications to image processing, computer vision, edge intelligence, and autonomous systems. The book is of relevance to a broad audience including researchers and practitioners active in the area of machine learning and pursuing fundamental and applied research in the area of advanced learning paradigms.

Comprehensive and up-to-date treatise of knowledge distillation cast in a general framework of transfer learning Focuses on a spectrum of methodological and algorithmic issues Includes recent developments in vision and language learning and relational architectures

Inhalt
Categories of Response-Based, Feature-Based, and Relation-Based Knowledge Distillation.- A Geometric Perspective on Feature-Based Distillation.- Knowledge Distillation Across Vision and Language.- Knowledge Distillation in Granular Fuzzy Models by Solving Fuzzy Relation Equations.- Ensemble Knowledge Distillation for Edge Intelligence in Medical Applications.- Self-Distillation with the New Paradigm in Multi-Task Learning.- Knowledge Distillation for Autonomous Intelligent Unmanned System.