paper-with-me

홈 › Papers

Hint-dynamic Knowledge Distillation

2022-11-30 · Yiyang Liu, Chenxin Li, Xiaotong Tu, Xinghao Ding, Yue Huang

Knowledge Distillation (KD) transfers the knowledge from a high-capacity teacher model to promote a smaller student model. Existing efforts guide the distillation by matching their prediction logits, feature embedding, etc., while leaving how to efficiently utilize them in junction less explored. In this paper, we propose Hint-dynamic Knowledge Distillation, dubbed HKD, which excavates the knowledge from the teacher' s hints in a dynamic scheme. The guidance effect from the knowledge hints usually varies in different instances and learning stages, which motivates us to customize a specific hint-learning manner for each instance adaptively. Specifically, a meta-weight network is introduced to generate the instance-wise weight coefficients about knowledge hints in the perception of the dynamical learning progress of the student model. We further present a weight ensembling strategy to eliminate the potential bias of coefficient estimation by exploiting the historical statics. Experiments on standard benchmarks of CIFAR-100 and Tiny-ImageNet manifest that the proposed HKD well boost the effect of knowledge distillation tasks.

📄 PDF Abstract BibTeX arXiv:2211.17059

Code (0)

등록된 구현이 없습니다.

Tasks

Knowledge Distillation

Methods 이 논문이 사용한 방법론

Knowledge Distillation A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions.…

Similar Papers 제목 키워드 기반

PURSUhInT: In Search of Informative Hint Points Based on Layer Clustering for Knowledge Distillation

2021-02-26 · Expert Systems with Applications, Elsevier 2022 11 · Reyhan Kevser Keser, Aydin Ayanzadeh, Omid Abdollahi Aghdam, Caglar Kilcioglu 외

One of the most efficient methods for model compression is hint distillation, where the student model is injected with information (hints) from several different layers of the teacher model. Although the selection of hin…

ClusteringKnowledge DistillationModel Compression

Knowledge Condensation Distillation

2022-07-12 · Chenxin Li, Mingbao Lin, Zhiyuan Ding, Nie Lin 외

Knowledge Distillation (KD) transfers the knowledge from a high-capacity teacher network to strengthen a smaller student. Existing methods focus on excavating the knowledge hints and transferring the whole knowledge to t…

Knowledge Distillation

Adaptive Multi-Teacher Multi-level Knowledge Distillation

2021-03-06 · Yuang Liu, Wei zhang, Jun Wang

Knowledge distillation~(KD) is an effective learning paradigm for improving the performance of lightweight student networks by utilizing additional supervision knowledge distilled from teacher networks. Most pioneering s…

Knowledge Distillation

Interactive Distillation for Cooperative Multi-Agent Reinforcement Learning

2026-01-08 · Minwoo Cho, Batuhan Altundas, Matthew Gombolay arxiv

Knowledge distillation (KD) has the potential to accelerate MARL by employing a centralized teacher for decentralized students but faces key bottlenecks. Specifically, there are (1) challenges in synthesizing high-perfor…

Multi-agent Reinforcement LearningKnowledge Distillation

One-for-All: Bridge the Gap Between Heterogeneous Architectures in Knowledge Distillation

2023-10-30 · NeurIPS 2023 11 · Zhiwei Hao, Jianyuan Guo, Kai Han, Yehui Tang 외

Knowledge distillation~(KD) has proven to be a highly effective approach for enhancing model performance through a teacher-student training scheme. However, most existing distillation methods are designed under the assum…

AllKnowledge Distillation