paper-with-me

Papers

Exploring Dark Knowledge under Various Teacher Capacities and Addressing Capacity Mismatch

2024-05-21 · Xin-Chun Li, Wen-Shu Fan, Bowen Tao, Le Gan, De-Chuan Zhan

Knowledge Distillation (KD) could transfer the `dark knowledge" of a well-performed yet large neural network to a weaker but lightweight one. From the view of output logits and softened probabilities, this paper goes deeper into the dark knowledge provided by teachers with different capacities. Two fundamental observations are: (1) a larger teacher tends to produce probability vectors that are less distinct between non-ground-truth classes; (2) teachers with different capacities are basically consistent in their cognition of relative class affinity. Abundant experimental studies verify these observations and in-depth empirical explanations are provided. The difference in dark knowledge leads to the peculiar phenomenon named `capacity mismatch" that a more accurate teacher does not necessarily perform as well as a smaller teacher when teaching the same student network. Enlarging the distinctness between non-ground-truth class probabilities for larger teachers could address the capacity mismatch problem. This paper explores multiple simple yet effective ways to achieve this goal and verify their success by comparing them with popular KD methods that solve the capacity mismatch.

📄 PDF Abstract BibTeX arXiv:2405.13078

Code (0)

등록된 구현이 없습니다.

Tasks

Knowledge Distillation

Similar Papers 제목 키워드 기반

Exploring Knowledge Distillation of a Deep Neural Network for Multi-Script identification

2021-02-20 · Shuvayan Ghosh Dastidar, Kalpita Dutta, Nibaran Das, Mahantapas Kundu 외

Multi-lingual script identification is a difficult task consisting of different language with complex backgrounds in scene text images. According to the current research scenario, deep neural networks are employed as tea…

Knowledge DistillationTransfer Learning

Adam: Dense Retrieval Distillation with Adaptive Dark Examples

2022-12-20 · Chongyang Tao, Chang Liu, Tao Shen, Can Xu 외

To improve the performance of the dual-encoder retriever, one effective approach is knowledge distillation from the cross-encoder ranker. Existing works construct the candidate passages following the supervised learning …

Knowledge DistillationRetrieval

Teaching to Teach by Structured Dark Knowledge

2018-09-27 · Ziliang Chen, Keze Wang, Liang Lin

To educate hyper deep learners, \emph{Curriculum Learnings} (CLs) require either human heuristic participation or self-deciding the difficulties of training instances. These coaching manners are blind to the coherent str…

Diversity

DARK: Diagonal-Anchored Repulsive Knowledge Distillation for Vision-Language Models under Extreme Compression

2026-03-05 · Numan Saeed, Asif Hanif, Fadillah Adamsyah Maani, Hussain Alasmawi 외 arxiv

Compressing vision-language models for on-device deployment is increasingly important in clinical settings, but knowledge distillation (KD) degrades sharply when the teacher-student capacity gap spans an order of magnitu…

Knowledge Distillation

Student Customized Knowledge Distillation: Bridging the Gap Between Student and Teacher

2021-01-01 · ICCV 2021 10 · Yichen Zhu, Yi Wang

Knowledge distillation (KD) transfers the dark knowledge from cumbersome networks (teacher) to lightweight (student) networks and expects the student to achieve more promising performance than training without the te…

image-classificationImage ClassificationKnowledge DistillationMulti-Task Learning+3