Research map / Label-efficient and robust learning / Knowledge distillation
Knowledge Distillation: A Survey
Jianping Gou, Baosheng Yu, Stephen J. Maybank, Dacheng Tao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| distill from a ResNet56 teacher with feature knowledge (FT) | ResNet20 · CIFAR10 | accuracy: 92.22 -> 93.15 | Improved |
| distill with relation knowledge (FSP) | ResNet8 · CIFAR10 | accuracy: 87.91 -> 88.70 | Improved |
| distill from ResNet152 with TOFD | ShuffleNetV2 · CIFAR100 | accuracy: 72.38 -> 76.68 | Improved |
| distill from ResNet152 with TOFD | ResNet50 · CIFAR100 | accuracy: 77.42 -> 84.74 | Improved |
| distill from ResNet50 with SSKD | MobileNetV2 · CIFAR100 | accuracy: 65.79 -> 72.57 | Improved |
| online distillation with Rocket-KD | WRN-16-1 · CIFAR100 | accuracy: 56.30 -> 67.00 | Improved |
| online mutual learning (DML) with MobileNet peer | ResNet32 · CIFAR100 | accuracy: 68.99 -> 76.13 | Improved |
| add online ensemble teacher (ONE) | ResNet32 · CIFAR10 | accuracy: 93.07 -> 94.01 | Improved |
| add attention-based online distillation (ACNs) | AlexNet · CIFAR100 | accuracy: 39.45 -> 46.27 | Improved |
| add self-distillation (SD) | ResNet32 · CIFAR100 | accuracy: 68.39 -> 71.29 | Improved |
| add Tf-KD self-distillation | ResNeXt29 · CIFAR100 | accuracy: 81.03 -> 82.08 | Improved |
| add self-distillation with response and feature knowledge | DenseNet40 · CIFAR10 | accuracy: 94.53 -> 94.80 | Improved |
Closest papers by meaning
- Apprentice: Using Knowledge Distillation Techniques To Improve Low-Precision Network Accuracy (ICLR 2018 · similarity 0.93)
- Knowledge Distillation and Student-Teacher Learning for Visual Intelligence: A Review and New Outlooks (TPAMI 2021 · similarity 0.93)
- Densely Guided Knowledge Distillation Using Multiple Teacher Assistants (ICCV 2021 · similarity 0.93)
- Self-Distillation: Towards Efficient and Compact Neural Networks (TPAMI 2021 · similarity 0.92)
- Improved Knowledge Distillation via Teacher Assistant (AAAI 2020 · similarity 0.92)
- Neural Collapse Inspired Knowledge Distillation (AAAI 2025 · similarity 0.90)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
