Research map / Detection and segmentation / Open-set and few-shot detection
Graph-Based High-Order Relation Discovery for Fine-Grained Recognition
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 |
|---|---|---|---|
| Use GaRD rather than the strongest listed baseline | Fine-grained recognition with ResNet-50 · CUB-200-2011 | accuracy: 88.5% -> 89.6% | Helped |
| Use GaRD rather than the strongest listed baseline | Fine-grained recognition with ResNet-50 · FGVC-Aircraft | accuracy: 93.8% -> 94.3% | Helped |
| Use GaRD rather than the strongest listed baseline | Fine-grained recognition with ResNet-50 · Stanford-Cars | accuracy: 94.8% -> 95.1% | Helped |
| Use GaRD rather than the strongest listed baseline | Fine-grained recognition with ResNet-50 · NA-birds | accuracy: 88.0% -> 88.0% | No clear effect |
| Add bilinear pooling (ablation) | ResNet-50 baseline · CUB-200-2011 | accuracy: 85.4% -> 87.0% | Helped |
| Add the relation-discovery module (ablation) | ResNet-50 baseline · CUB-200-2011 | accuracy: 85.4% -> 88.1% | Helped |
| Add group-wise learning to relation discovery (ablation) | ResNet-50 · CUB-200-2011 | accuracy: 88.1% -> 88.8% | Helped |
| Add graph-based semantic grouping (ablation) | Relation discovery with group-wise learning · CUB-200-2011 | accuracy: 88.8% -> 89.6% | Helped |
| Remove spatial enhancement from relation discovery (ablation) | GaRD feature extraction · CUB-200-2011 | accuracy: 88.6% -> 88.1% | Hurt |
| Remove multi-level aggregation from relation discovery (ablation) | GaRD feature extraction · CUB-200-2011 | accuracy: 88.6% -> 88.3% | Hurt |
| Replace graph embedding with mean aggregation (ablation) | GaRD · CUB-200-2011 | accuracy: 89.6% -> 88.6% | Hurt |
| Replace learned graph embedding with a static all-ones adjacency matrix (ablation) | GaRD · CUB-200-2011 | accuracy: 89.6% -> 88.5% | Hurt |
Closest papers by meaning
- Spatial-Aware Graph Relation Network for Large-Scale Object Detection (CVPR 2019 · similarity 0.88)
- Graph-Based High-Order Relation Modeling for Long-Term Action Recognition (CVPR 2021 · similarity 0.87)
- Interpretable and Accurate Fine-grained Recognition via Region Grouping (CVPR 2020 · similarity 0.86)
- GUIDED: Granular Understanding via Identification, Detection, and Discrimination for Fine-Grained Open-Vocabulary Object Detection (NeurIPS 2025 · similarity 0.85)
- Democratizing Fine-grained Visual Recognition with Large Language Models (ICLR 2024 · similarity 0.84)
- Multi-Part Object Representations via Graph Structures and Co-Part Discovery (NeurIPS 2026 · 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.