Research map / Graph learning / Graph self-supervised learning
Towards Human-Like Robot Handwriting via Contour-Aware Generation
Yutao Qin, Gang Dai, Yifan Zhang, Youwei Han, Qisheng He, Shuangping Huang
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 contour-aware graph encoder to the base model (ablation) | G-HTR base · CHTR-110K test (All) | mIOU: 0.532 -> 0.596 | Improved |
| add graph encoder to the base model (ablation) | G-HTR base · CHTR-110K test (All) | DTW: 19.228 -> 15.512 | Improved |
| add multi-scale graph learning on top of graph encoder (ablation) | G-HTR · CHTR-110K test (All) | mIOU: 0.596 -> 0.641 | Improved |
| add multi-scale graph learning on top of graph encoder (ablation) | G-HTR · CHTR-110K test (All) | DTW: 15.512 -> 12.765 | Improved |
| add multi-scale graph learning on top of graph encoder (ablation) | G-HTR · CHTR-110K test (All) | FID: 1.346 -> 1.228 | Improved |
| adapt baseline input embedding and output projection to model stroke width (ablation) | TrajFormer · CHTR-110K test (All) | mIOU: 0.104 -> 0.552 | Improved |
| adapt baseline to model stroke width (ablation) | PEN-Net · CHTR-110K test (All) | mIOU: 0.048 -> 0.144 | Improved |
| G-HTR vs TrajFormer on contour-aware trajectory reconstruction | G-HTR · CHTR-110K test (All) | mIOU: 0.552 -> 0.641 | Improved |
| G-HTR vs TrajFormer on conventional trajectory recovery | G-HTR · CASIA-OLHWDB | DTW: 23.892 -> 16.264 | Improved |
| G-HTR vs TrajFormer on CASIA-OLHWDB FID | G-HTR · CASIA-OLHWDB | FID: 1.121 -> 1.050 | Improved |
| G-HTR on rare characters outside GB2312-80 | G-HTR · failure analysis | stroke order accuracy | Hurt |
Closest papers by meaning
- Bridging Online and Offline Handwriting via Differentiable Physical Rendering (ECCV 2026 · similarity 0.88)
- BézierSketch: A generative model for scalable vector sketches (ECCV 2020 · similarity 0.85)
- Disentangling Writer and Character Styles for Handwriting Generation (CVPR 2023 · similarity 0.84)
- StrokeFusion: Vector Sketch Generation via Joint Stroke-UDF Encoding and Latent Sequence Diffusion (AAAI 2026 · similarity 0.84)
- DeepWriteSYN: On-Line Handwriting Synthesis via Deep Short-Term Representations (AAAI 2021 · similarity 0.84)
- SpaceGTN: A Time-Agnostic Graph Transformer Network for Handwritten Diagram Recognition and Segmentation (AAAI 2024 · similarity 0.83)
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.
