Research map / Graph learning / Knowledge and temporal graphs
End-to-End Hyper-Relational Information Extraction for Engineering Diagrams via Dynamically Tokenized Relation Transformer
Tianyou Bai, Yan-Ming Zhang, Zixiang Zhang, Jibin Zhou, Fei Yin, Cheng-Lin Liu
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 denoising (DN) training (ablation) | relation Transformer · P&ID | AP50: 90.12 -> 93.94 | Improved |
| add relation-aware deformable decoupled visual attention (DDVA) (ablation) | relation Transformer with DN · P&ID | R@1000 (SGDET): 70.49 -> 77.65 | Improved |
| add contrastive denoising (CDN) (ablation) | relation Transformer with DN and DDVA · P&ID | R@1000 (SGDET): 77.65 -> 81.96 | Improved |
| add mixed query selection (MQS) (ablation) | relation Transformer with DN, DDVA, CDN · P&ID | R@1000 (SGDET): 81.96 -> 87.42 | Improved |
| add token scorers to prune 70% of tokens (ablation) | Swin backbone · P&ID | GFLOPs: 299.2 -> 90.5 | Improved |
| add token scorers only (no reconstruction branch) (ablation) | Swin backbone · P&ID | R@1000 (SGCLS): 87.99 -> 70.12 | Hurt |
| add reconstruction branch on top of scorers (ablation) | token pruning · P&ID | R@1000 (SGCLS): 70.12 -> 87.42 | Improved |
| add scorers only (ablation) | Swin backbone · OCR | CER: 1.21 -> 9.86 | Hurt |
| add reconstruction branch on top of scorers (ablation) | token pruning · OCR | CER: 9.86 -> 1.53 | Improved |
| replace two-stage Relationformer with one-stage DTRT | scene graph generation · P&ID | GFLOPs: 791.9 -> 90.5 | Improved |
| replace Relationformer with DTRT | scene graph detection · P&ID | R@1000: 76.43 -> 87.42 | Improved |
| replace RT-DETR with DTRT for entity detection | P&ID symbol detection · P&ID | AP50: 99.27 -> 99.01 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- Relationformer — Relationformer: A Unified Framework for Image-to-Graph Generation (ECCV 2022) · better on 8 of 9 shared results
- RelTR — RelTR: Relation Transformer for Scene Graph Generation (TPAMI 2023) · better on 2 of 2 shared results
- SGTR — SGTR: End-to-End Scene Graph Generation With Transformer (CVPR 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- From Sequence to Structure: Uncovering Substructure Reasoning in Transformers (NeurIPS 2025 · similarity 0.83)
- Text2NKG: Fine-Grained N-ary Relation Extraction for N-ary relational Knowledge Graph Construction (NeurIPS 2024 · similarity 0.83)
- CircuitSense: A Hierarchical Circuit System Benchmark Bridging Visual Comprehension and Symbolic Reasoning in Engineering Design Process (ICLR 2026 · similarity 0.82)
- Efficient Automated Circuit Discovery in Transformers using Contextual Decomposition (ICLR 2025 · similarity 0.81)
- Structure Is All You Need: Structural Representation Learning on Hyper-Relational Knowledge Graphs (ICML 2025 · similarity 0.80)
- MarkushGrapher: Joint Visual and Textual Recognition of Markush Structures (CVPR 2025 · similarity 0.80)
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.
