Research map / Detection and segmentation / CNN features
Semantic-Aligned Fusion Transformer for One-Shot 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 |
|---|---|---|---|
| Add semantic-aligned VFM and HFM fusion | FCOS one-shot detection · PASCAL-VOC 2007 test, Novel Avg. | AP50: 74.5 vs 73.1 | Helped |
| Add semantic-aligned VFM and HFM fusion | FCOS one-shot detection · MS-COCO 2017 val, Novel Avg. | AP50: 24.9 vs 24.3 | Helped |
| Add HFM instead of reweighting or correlation (ablation) | VOC detector with VFM · PASCAL-VOC 2007 test, Novel | AP50: 64.2 -> 71.7; 64.3 -> 71.7 | Helped |
| Replace FPN with VFM (ablation) | VOC detector with HFM · PASCAL-VOC 2007 test, Novel | AP50: 69.2 -> 71.7 | Helped |
| Use corresponding-scale rather than one-to-all-scale fusion (ablation) | SaFT with VFM and HFM · PASCAL-VOC 2007 test, Novel | AP50: 69.9 -> 71.7 | Hurt |
| Add level 5 to corresponding-scale feature fusion (ablation) | SaFT · PASCAL-VOC 2007 test, Novel | AP50: 61.4 -> 70.0 | Helped |
| Add level 6 to corresponding-scale feature fusion (ablation) | SaFT · PASCAL-VOC 2007 test, Novel | AP50: 70.0 -> 69.9 | Hurt |
| Use level 5 rather than level 4 as the only support feature (ablation) | SaFT one-to-all-scale fusion · PASCAL-VOC 2007 test, Novel | AP50: 68.1 -> 71.7 | Helped |
| Increase HA blocks from 2 to 4 (ablation) | SaFT · PASCAL-VOC 2007 test, Novel | AP50: 69.5 -> 71.7 | Helped |
| Increase HA blocks from 4 to 6 (ablation) | SaFT · PASCAL-VOC 2007 test, Novel | AP50: 71.7 -> 72.8 | Helped |
Closest papers by meaning
- Adaptive Image Transformer for One-Shot Object Detection (CVPR 2021 · similarity 0.89)
- Semantic-Aware Transformation-Invariant RoI Align (AAAI 2024 · similarity 0.88)
- Inference Fusion with Associative Semantics for Unseen Object Detection (AAAI 2021 · similarity 0.88)
- Accelerating DETR Convergence via Semantic-Aligned Matching (CVPR 2022 · similarity 0.88)
- SMR-Net: Semantic-Guided Mutually Reinforcing Network for Cross-Modal Image Fusion and Salient Object Detection (AAAI 2025 · similarity 0.88)
- Fine-Grained Prototypes Distillation for Few-Shot Object Detection (AAAI 2024 · 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.