paper-with-me

홈 › Papers

Do Not Blindly Imitate the Teacher: Using Perturbed Loss for Knowledge Distillation

2023-05-08 · Rongzhi Zhang, Jiaming Shen, Tianqi Liu, Jialu Liu, Michael Bendersky, Marc Najork, Chao Zhang

Knowledge distillation is a popular technique to transfer knowledge from large teacher models to a small student model. Typically, the student learns to imitate the teacher by minimizing the KL divergence of its output distribution with the teacher's output distribution. In this work, we argue that such a learning objective is sub-optimal because there exists a discrepancy between the teacher's output distribution and the ground truth label distribution. Therefore, forcing the student to blindly imitate the unreliable teacher output distribution leads to inferior performance. To this end, we propose a novel knowledge distillation objective PTLoss by first representing the vanilla KL-based distillation loss function via a Maclaurin series and then perturbing the leading-order terms in this series. This perturbed loss implicitly transforms the original teacher into a proxy teacher with a distribution closer to the ground truth distribution. We establish the theoretical connection between this "distribution closeness" and the student model generalizability, which enables us to select the PTLoss's perturbation coefficients in a principled way. Extensive experiments on five datasets demonstrate PTLoss can significantly improve the distillation effectiveness for teachers of various scales.

📄 PDF Abstract BibTeX arXiv:2305.05010

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

Towards Oracle Knowledge Distillation with Neural Architecture Search

2019-11-29 · Minsoo Kang, Jonghwan Mun, Bohyung Han

We present a novel framework of knowledge distillation that is capable of learning powerful and efficient student models from ensemble teacher networks. Our approach addresses the inherent model capacity issue between te…

image-classificationImage ClassificationKnowledge DistillationNeural Architecture Search

Relative Difficulty Distillation for Semantic Segmentation

2024-07-04 · Dong Liang, Yue Sun, Yun Du, Songcan Chen 외

Current knowledge distillation (KD) methods primarily focus on transferring various structured knowledge and designing corresponding optimization goals to encourage the student network to imitate the output of the teache…

Knowledge DistillationSemantic Segmentation

KD-Zero: Evolving Knowledge Distiller for Any Teacher-Student Pairs

2023-09-21 · NeurIPS 2023 11

Knowledge distillation (KD) has emerged as an effective technique for compressing models that can enhance the lightweight model. Conventional KD methods propose various designs to allow student model to imitate the teac…

Rethinking the Knowledge Distillation From the Perspective of Model Calibration

2021-10-31 · Lehan Yang, Jincen Song

Recent years have witnessed dramatically improvements in the knowledge distillation, which can generate a compact student model for better efficiency while retaining the model effectiveness of the teacher model. Previous…

Knowledge Distillation

Differentiable Feature Aggregation Search for Knowledge Distillation

2020-08-02 · ECCV 2020 8 · Yushuo Guan, Pengyu Zhao, Bingxuan Wang, Yuanxing Zhang 외

Knowledge distillation has become increasingly important in model compression. It boosts the performance of a miniaturized student network with the supervision of the output distribution and feature maps from a sophistic…

Knowledge DistillationModel CompressionNeural Architecture Search