Research map / Detection and segmentation / CNN features
Dynamic Anchor Feature Selection for Single-Shot Object Detection
Shuai Li, Lingxiao Yang, Jianqiang Huang, Xian-Sheng Hua, Lei Zhang
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 |
|---|---|---|---|
| add DAFS feature selection instead of RoIPooling-style anchor pooling (ablation) | DAFS320 · VOC2007 test | AP: 50.9 -> 57.0 | Improved |
| add DAFS feature selection instead of deformable convolution (ablation) | DAFS320 · VOC2007 test | AP: 53.6 -> 57.0 | Improved |
| add BFF transfer block between ARM and ODM instead of none (ablation) | DAFS320 · VOC2007 test | AP: 56.0 -> 57.0 | Improved |
| use BFF instead of TCB (ablation) | DAFS320 · VOC2007 test | AP: 56.6 -> 57.0 | Improved |
| use class-specific instead of class-agnostic ARM classifier (ablation) | DAFS320 · VOC2007 test | AP: 56.6 -> 57.0 | Improved |
| use class-specific instead of class-agnostic ARM classifier (ablation) | DAFS320 · VOC2007 test | AP80: 46.1 -> 46.0 | No clear effect |
| increase default anchors from 1 to 5 per pixel (ablation) | DAFS320 · VOC2007 test | AP90: 17.3 -> 20.4 | Improved |
| increase default anchors from 1 to 5 per pixel (ablation) | DAFS320 · VOC2007 test | AP50: 80.6 -> 80.3 | No clear effect |
| increase feature sampling points from (3,3,6,6) to (9,9,9,9) (ablation) | DAFS320 · VOC2007 test | AP: 57.0 -> 56.7 | No clear effect |
| add DAFS to RefineDet320 (VGG16) | RefineDet320 · COCO test-dev | AP: 29.4 -> 31.2 | Improved |
| add DAFS to RefineDet320 (VGG16) | RefineDet320 · VOC2007 test (COCO-style) | AP90: 12.2 -> 17.3 | Improved |
| add DAFS to SSD | SSD320 · VOC2007 test | AP: 47.1 -> 49.8 | Improved |
Closest papers by meaning
- Dynamic Anchor Learning for Arbitrary-Oriented Object Detection (AAAI 2021 · similarity 0.89)
- Soft Anchor-Point Object Detection (ECCV 2020 · similarity 0.89)
- Feature Selective Anchor-Free Module for Single-Shot Object Detection (CVPR 2019 · similarity 0.89)
- Bridging the Gap Between Anchor-Based and Anchor-Free Detection via Adaptive Training Sample Selection (CVPR 2020 · similarity 0.89)
- Multiple Anchor Learning for Visual Object Detection (CVPR 2020 · similarity 0.87)
- Probabilistic Anchor Assignment with IoU Prediction for Object Detection (ECCV 2020 · similarity 0.87)
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.
