Research map / Detection and segmentation / CNN features
OPANAS: One-Shot Path Aggregation Network Architecture 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 |
|---|---|---|---|
| Replace the baseline FPN with OPA-FPN | RetinaNet · COCO minival | mAP: 35.7 -> 38.0 | Helped |
| Replace the baseline FPN with OPA-FPN | Faster R-CNN · COCO minival | mAP: 36.4 -> 39.6 | Helped |
| Replace the baseline FPN with OPA-FPN | Cascade R-CNN · COCO minival | mAP: 40.3 -> 42.8 | Helped |
| Replace a single information path with searched aggregation of multiple paths | Faster R-CNN, ResNet50 · COCO minival | mAP: 36.4 -> 39.6 | Helped |
| Use only the bottom-up information path instead of the baseline FPN (ablation) | Faster R-CNN, ResNet50 · COCO minival | mAP: 36.4 -> 30.0 | Hurt |
| Use only the scale-equalizing information path instead of the baseline FPN (ablation) | Faster R-CNN, ResNet50 · COCO minival | mAP: 36.4 -> 32.6 | Hurt |
| Use only the top-down information path instead of the baseline FPN (ablation) | Faster R-CNN, ResNet50 · COCO minival | mAP: 36.4 -> 37.5 | Helped |
| Use only the fusing-splitting information path instead of the baseline FPN (ablation) | Faster R-CNN, ResNet50 · COCO minival | mAP: 36.4 -> 37.6 | Helped |
| Add fair sampling to the densely connected super-net (ablation) | OPANAS · COCO minival | Kendall tau: 0.3584 -> 0.5865 | Helped |
| Add edge importance weights to the densely connected super-net with fair sampling (ablation) | OPANAS · COCO minival | Kendall tau: 0.5865 -> 0.6145 | Helped |
| Use OPANAS search instead of alternative NAS search methods | Faster R-CNN, ResNet50 · COCO minival | mAP: SPOS 38.4; DARTS 39.1; FairDARTS 39.4; OPANAS 39.6 | Helped |
Closest papers by meaning
- NAS-FPN: Learning Scalable Feature Pyramid Architecture for Object Detection (CVPR 2019 · similarity 0.89)
- DetNAS: Backbone Search for Object Detection (NeurIPS 2019 · similarity 0.89)
- NAS-FCOS: Fast Neural Architecture Search for Object Detection (CVPR 2020 · similarity 0.89)
- SM-NAS: Structural-to-Modular Neural Architecture Search for Object Detection (AAAI 2020 · similarity 0.88)
- Feature Pyramid Networks for Object Detection (CVPR 2017 · similarity 0.87)
- DetKDS: Knowledge Distillation Search for Object Detectors (ICML 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.