Research map / Architectures and efficiency / CNN design
LCNN: Lookup-Based Convolutional Neural Network
Check 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 the smaller-dictionary LCNN-fast configuration instead of AlexNet CNN | AlexNet · ILSVRC2012 val | speedup: 1.0× -> 37.6× | Helped |
| Use the smaller-dictionary LCNN-fast configuration instead of AlexNet CNN | AlexNet · ILSVRC2012 val | top-1 accuracy: 56.6 -> 44.3 | Hurt |
| Use the larger-dictionary LCNN-accurate configuration instead of AlexNet CNN | AlexNet · ILSVRC2012 val | top-1 accuracy: 56.6 -> 55.1 | Hurt |
| Increase dictionary size for the accurate configuration rather than the fast configuration | AlexNet LCNN · ILSVRC2012 val | top-1 accuracy: 44.3 -> 55.1 | Helped |
| Increase dictionary size for the accurate configuration rather than the fast configuration | AlexNet LCNN · ILSVRC2012 val | speedup: 37.6× -> 3.2× | Hurt |
| Use LCNN-fast instead of ResNet-18 CNN | ResNet-18 · ILSVRC2012 val | speedup: 1.0× -> 29.2× | Helped |
| Use LCNN-fast instead of ResNet-18 CNN | ResNet-18 · ILSVRC2012 val | top-1 accuracy: 69.3 -> 51.8 | Hurt |
| Use LCNN-accurate instead of ResNet-18 CNN | ResNet-18 · ILSVRC2012 val | top-1 accuracy: 69.3 -> 62.2 | Hurt |
| Transfer dictionaries from a 10-layer ResNet and train only lookup parameters in ResNet-18 | LCNN ResNet-18 · ImageNet, 10K iterations | top-1 accuracy: 16.2% higher | Helped |
| Keep the pretrained dictionaries fixed and fine-tune sparse parameters for novel classes | AlexNet LCNN · ILSVRC2012 few-shot learning, 1 example per category | top-1 accuracy: +6.3% | Helped |
Closest papers by meaning
- More Is Less: A More Complicated Network With Less Inference Complexity (CVPR 2017 · similarity 0.88)
- Doubly Convolutional Neural Networks (NeurIPS 2016 · similarity 0.87)
- LSNet: See Large, Focus Small (CVPR 2025 · similarity 0.86)
- EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks (ICML 2019 · similarity 0.85)
- KernelWarehouse: Rethinking the Design of Dynamic Convolution (ICML 2024 · similarity 0.84)
- LAuReL: Learned Augmented Residual Layer (ICML 2025 · similarity 0.84)
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.