Research map / Architectures and efficiency / Neural architecture search
One-Shot Neural Ensemble Architecture Search by Diversity-Guided Search Space Shrinking
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 |
|---|---|---|---|
| add diversity-guided search space shrinking instead of no shrinking (ablation) | NEAS · ImageNet | Top-1 accuracy (%): 77.3 -> 77.9 | Helped |
| replace accuracy-based shrinking with diversity-score shrinking (ablation) | NEAS · ImageNet | Kendall Tau: 0.42 -> 0.65 | Helped |
| replace accuracy-based shrinking with diversity-score shrinking (ablation) | NEAS · ImageNet | Top-1 accuracy (%): 77.2 -> 77.9 | Helped |
| add diversity-guided shrinking rather than no shrinking (ablation) | NEAS supernet · ImageNet, supernet-inherited weights | Top-1 accuracy (%): 67.8 -> 68.3 | Helped |
| use heterogeneous paths instead of two identical paths (ablation) | two-path ensemble · ImageNet | Top-1 accuracy (%): 79.4 -> 80.0; 79.6 -> 80.0 | Helped |
| add layer sharing (ablation) | NEAS ensemble · ImageNet | FLOPs (M): 605 -> 574 | Helped |
| increase ensemble paths from 2 to 3 (ablation) | NEAS · ImageNet | Top-1 accuracy (%): 80.0 -> 79.5 | Hurt |
| increase ensemble paths from 2 to 5 (ablation) | NEAS · ImageNet | Top-1 accuracy (%): 80.0 -> 78.5 | Hurt |
| search heterogeneous ensemble architectures with NEAS rather than use MobileNetV3 | COCO backbone · COCO val2017 | AP (%): 29.9 -> 33.0 | Helped |
Closest papers by meaning
- Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective (ICLR 2021 · similarity 0.90)
- Efficient Few-Shot Neural Architecture Search by Counting the Number of Nonlinear Functions (AAAI 2025 · similarity 0.90)
- einspace: Searching for Neural Architectures from Fundamental Operations (NeurIPS 2024 · similarity 0.89)
- Vision-Oriented Lightweight Neural Architecture Search with Budget-Adaptive Evaluation (CVPR 2026 · similarity 0.89)
- Understanding Architectures Learnt by Cell-based Neural Architecture Search (ICLR 2020 · similarity 0.89)
- Boosting Order-Preserving and Transferability for Neural Architecture Search: a Joint Architecture Refined Search and Fine-tuning Approach (CVPR 2024 · similarity 0.88)
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.