Research map / Architectures and efficiency / CNN design
Multigrid Neural Architectures
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 |
|---|---|---|---|
| Replace a single grid with multigrid in VGG-11 | CIFAR-100 classification · CIFAR-100 | Error (%): 30.26 -> 28.39 | Helped |
| Add multigrid to residual CNNs | CIFAR-100 classification · CIFAR-100 | Error (%): 28.05 -> 26.79; 27.12 -> 25.29 | Helped |
| Process progressively from coarse to full pyramid | CIFAR-100 classification · CIFAR-100 | Error (%): 30.26 -> 28.11 | Helped |
| Reduce channel counts in MG-16 (ablation) | Multigrid VGG-16 · CIFAR-100 | Error (%): 29.91 -> 32.59 | Hurt |
| Reduce channel counts in PMG-16 (ablation) | Progressive multigrid VGG-16 · CIFAR-100 | Error (%): 28.11 -> 30.47 | Hurt |
| Reduce channel counts in the residual multigrid model (ablation) | CIFAR-100 classification · CIFAR-100 | Error (%): 25.29; 26.85 | Mixed |
| Use progressive multigrid for segmentation instead of a single-grid model | MNIST semantic segmentation · MNIST semantic segmentation | mean IoU (%): 32.21 -> 61.88 | Helped |
| Add progressive multigrid to the residual segmentation model | MNIST semantic segmentation · MNIST semantic segmentation | mean IoU (%): 69.86 -> 81.91 | Helped |
| Use progressive multigrid for spatial transformation | MNIST spatial transformation · MNIST spatial transformation | mean IoU (%): 12.20 -> 50.64 | Helped |
| Add progressive multigrid to the residual transformation model | MNIST spatial transformation · MNIST spatial transformation | mean IoU (%): 0 -> 55.61 | Helped |
| Add multigrid to ResNet-34 | ImageNet classification · ImageNet validation, 10-crop | Top-1 error (%): 24.19 -> 22.42 | Helped |
| Replace a wider residual network with multigrid | ImageNet classification · ImageNet validation, single-crop | Top-1 error (%): 24.50 -> 24.51 | No clear effect |
Closest papers by meaning
- Multigrid Neural Memory (ICML 2020 · similarity 0.93)
- DiffGCN: Graph Convolutional Networks via Differential Operators and Algebraic Multigrid Pooling (NeurIPS 2020 · similarity 0.87)
- Unveiling the Unseen: Identifiable Clusters in Trained Depthwise Convolutional Kernels (ICLR 2024 · similarity 0.85)
- PICNN: A Pathway towards Interpretable Convolutional Neural Networks (AAAI 2024 · similarity 0.84)
- Convolutions and More as Einsum: A Tensor Network Perspective with Advances for Second-Order Methods (NeurIPS 2024 · similarity 0.84)
- Multiscale Low-Frequency Memory Network for Improved Feature Extraction in Convolutional Neural Networks (AAAI 2024 · 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.