方法对比

	胶囊网络	知识蒸馏	神经架构搜索
领域	深度学习	深度学习	深度学习
方法族	Machine learning	Machine learning	Machine learning
起源年份	2017	2015	2017
提出者	Sabour, S., Frosst, N. & Hinton, G. E.	Hinton, G., Vinyals, O. & Dean, J.	Zoph, B. & Le, Q.V.
类型	Deep learning architecture (vector capsules with dynamic routing)	Neural network compression (teacher–student)	Automated architecture optimization (deep learning)
开创性文献	Sabour, S., Frosst, N. & Hinton, G. E. (2017). Dynamic Routing Between Capsules. Advances in Neural Information Processing Systems (NeurIPS).	Hinton, G., Vinyals, O. & Dean, J. (2015). Distilling the Knowledge in a Neural Network. NeurIPS Deep Learning Workshop.	Zoph, B. & Le, Q.V. (2017). Neural Architecture Search with Reinforcement Learning. ICLR.
别名	Kapsül Ağı (CapsNet), CapsNet, capsule net, dynamic routing network	Bilgi Damıtma (Knowledge Distillation), bilgi damıtma, teacher-student distillation, model distillation	Nöral Mimari Arama (NAS), NAS, automated architecture design, differentiable architecture search
相关	4	5	5
摘要	A Capsule Network (CapsNet) is a deep learning architecture introduced by Sara Sabour, Nicholas Frosst and Geoffrey Hinton in 2017 that organises neurons as vectors (capsules) rather than scalar activations, so that spatial hierarchy and pose (orientation) information are encoded directly. It was proposed to overcome the fragility of convolutional networks to changes in viewpoint.	Knowledge Distillation is a model-compression technique, introduced by Geoffrey Hinton and colleagues in 2015, that trains a small student model using the soft-label outputs of a large teacher model. Distilled models such as DistilBERT and TinyBERT reach roughly 97% of the larger model's performance while running far faster.	Neural Architecture Search (NAS), introduced by Zoph and Le in 2017, automatically optimizes architectural decisions such as a network's depth, width, and connection structure instead of hand-designing them. Leading methods in the field include DARTS, ENAS, and Once-for-All.