Research map / Detection and segmentation / CNN features
SP-NAS: Serial-to-Parallel Backbone Search for Object Detection
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 |
|---|---|---|---|
| Search a task-specific backbone with SP-NAS | Cascade RCNN with SPNetCOCO (XB) · COCO | AP: 47.4 | Helped |
| Search a task-specific backbone with SP-NAS | SPNetECP · ECP | LAMR: 0.067 -> 0.054 | Helped |
| Search a task-specific backbone with SP-NAS | SPNetBDD · BDD | AP: 37.1 -> 38.7 | Helped |
| Search a task-specific backbone with SP-NAS | SPNetVOC · VOC | AP50: 81.8 -> 84.1 | Helped |
| Add the first serial-level search round (ablation) | SP-NAS on BDD · BDD | mAP: 36.5 -> 37.2+0.7 | Helped |
| Add the second serial-level search round (ablation) | SP-NAS on BDD · BDD | mAP: 37.2+0.7 -> 38.0+0.8 | Helped |
| Add parallel-level search after serial-level search (ablation) | SP-NAS on BDD · BDD | mAP: 38.0+0.8 -> 38.7+0.7 | Helped |
| Add the first serial-level search round (ablation) | SP-NAS on ECP · ECP | LAMR: 0.088 -> 0.064−0.024 | Helped |
| Add the second serial-level search round (ablation) | SP-NAS on ECP · ECP | LAMR: 0.064−0.024 -> 0.061−0.003 | Helped |
| Add parallel-level search after serial-level search (ablation) | SP-NAS on ECP · ECP | LAMR: 0.061−0.003 -> 0.054−0.007 | Helped |
| Add the first serial-level search round (ablation) | SP-NAS on VOC · VOC | AP50: 77.1 -> 81.2+4.1 | Helped |
| Add the second serial-level search round (ablation) | SP-NAS on VOC · VOC | AP50: 81.2+4.1 -> 83.5+2.3 | Helped |
Closest papers by meaning
- SM-NAS: Structural-to-Modular Neural Architecture Search for Object Detection (AAAI 2020 · similarity 0.94)
- NAS-FCOS: Fast Neural Architecture Search for Object Detection (CVPR 2020 · similarity 0.92)
- DetNAS: Backbone Search for Object Detection (NeurIPS 2019 · similarity 0.91)
- Hit-Detector: Hierarchical Trinity Architecture Search for Object Detection (CVPR 2020 · similarity 0.90)
- NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection (CVPR 2019 · similarity 0.90)
- Joint-DetNAS: Upgrade Your Detector With NAS, Pruning and Dynamic Distillation (CVPR 2021 · similarity 0.89)
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.