Research map / Architectures and efficiency / CNN design
Kernel Based Progressive Distillation for Adder Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use a progressively learned teacher instead of a fixed pretrained teacher (ablation) | PKKD ANN-20 · CIFAR-10 | accuracy (%): 92.39 -> 92.96 | Helped |
| add kernel transformation to feature KD (with 1x1 conv) (ablation) | PKKD ANN-20 · CIFAR-10 | accuracy (%): 92.67 -> 92.96 | Helped |
| add 1x1 conv after kernel (ablation) | PKKD ANN-20 · CIFAR-10 | accuracy (%): 92.75 -> 92.96 | Helped |
| add 1x1 conv to plain KD loss (ablation) | ANN-20 KD · CIFAR-10 | accuracy (%): 92.58 -> 92.67 | Helped |
| add kernel to plain KD loss (ablation) | ANN-20 KD · CIFAR-10 | accuracy (%): 92.58 -> 92.75 | Helped |
| increase alpha=beta from 0.1 to 10 (ablation) | PKKD ANN-20 · CIFAR-10 | accuracy (%): 92.49 -> 92.96 | Helped |
| increase alpha=beta from 0.1 to 10 (ablation) | PKKD ANN-20 · CIFAR-100 | accuracy (%): 68.51 -> 69.93 (best at 5), 69.62 at 10 | Mixed |
| apply PKKD to ANN instead of vanilla training | ANN ResNet-50 · ImageNet | top-1 accuracy (%): 74.9 -> 76.8 | Helped |
| apply PKKD to ANN instead of vanilla training | ANN ResNet-20 · CIFAR-100 | accuracy (%): 67.60 -> 69.93 | Helped |
Closest papers by meaning
- Towards Stable and Robust AdderNets (NeurIPS 2021 · similarity 0.85)
- VkD: Improving Knowledge Distillation using Orthogonal Projections (CVPR 2024 · similarity 0.85)
- Data Distillation Can Be Like Vodka: Distilling More Times For Better Quality (ICLR 2024 · similarity 0.84)
- Progressively Knowledge Distillation via Re-parameterizing Diffusion Reverse Process (AAAI 2024 · similarity 0.84)
- Provable and Efficient Dataset Distillation for Kernel Ridge Regression (NeurIPS 2024 · similarity 0.83)
- Dataset Distillation via Knowledge Distillation: Towards Efficient Self-Supervised Pre-training of Deep Networks (ICLR 2025 · similarity 0.83)
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.