Research map / Architectures and efficiency / Neural architecture search
Genetic CNN
Lingxi Xie, Alan Yuille
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 |
|---|---|---|---|
| run 50 generations of genetic search instead of the initial random population | GeNet · CIFAR10 test | error rate (%): error 24.04% -> 22.81% | Improved |
| replace random initialization with all-zero initialization (ablation) | GeNet · CIFAR10 test | accuracy (%) | No clear effect |
| use genetic search instead of random search with same budget of 1020 individuals (ablation) | GeNet · CIFAR10 test | accuracy (%): 76.94 -> 77.19 | Improved |
| use the searched structure instead of the densely connected all-ones string (ablation) | GeNet · CIFAR10 test | accuracy (%): 76.84 -> 77.19 | Improved |
| run longer genetic search (G-00 to G-50), transferred to CIFAR10 (ablation) | GeNet · CIFAR10 error (Table 2) | error rate (%): 8.18 -> 7.19 | Improved |
| run longer genetic search (G-00 to G-50), transferred to CIFAR100 (ablation) | GeNet · CIFAR100 error (Table 2) | error rate (%): 31.46 -> 29.03 | Improved |
| run longer genetic search (G-00 to G-50), transferred to SVHN (ablation) | GeNet · SVHN error (Table 2) | error rate (%): 2.25 -> 1.99 | Improved |
| replace VGGNet-16 chain stages with GeNet #2 structure | VGGNet-16 · ILSVRC2012 val top-1 | top-1 error (%): 28.5 -> 27.87 | Improved |
| replace VGGNet-16 chain stages with GeNet #2 structure | VGGNet-16 · ILSVRC2012 val top-5 | top-5 error (%): 9.9 -> 9.74 | Improved |
| replace VGGNet-16 chain stages with GeNet #1 structure | VGGNet-16 · ILSVRC2012 val top-5 | top-5 error (%): 9.9 -> 9.95 | No clear effect |
| start genetic search from the 40-layer wide residual network | WRN-40 · CIFAR10 error | error rate (%): 5.54 -> 5.39 | Improved |
| replace VGGNet-16 features with GeNet #2 features | VGGNet-16 · Caltech256 | accuracy (%): 82.69 -> 83.78 | Improved |
Closest papers by meaning
- Bilinear Convolutional Neural Networks for Fine-Grained Visual Recognition (TPAMI 2017 · similarity 0.87)
- Hierarchical Representations for Efficient Architecture Search (ICLR 2018 · similarity 0.86)
- Involution: Inverting the Inherence of Convolution for Visual Recognition (CVPR 2021 · similarity 0.86)
- Linear CNNs Discover the Statistical Structure of the Dataset Using Only the Most Dominant Frequencies (ICML 2023 · similarity 0.86)
- Accelerating Very Deep Convolutional Networks for Classification and Detection (TPAMI 2016 · similarity 0.86)
- Neural Genetic Search in Discrete Spaces (ICML 2025 · similarity 0.85)
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.
