Research map / Detection and segmentation / CNN features
Fully Convolutional Network Based Skeletonization for Handwritten Chinese Characters
Read the paperCheck 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 learnable upsampling in place of bilinear (ablation) | baseline FCN · Skeleton extraction, ICDAR-2013 synthesized | F-measure: 0.575 -> 0.592 | Helped |
| add multi-rate dilated fusion (MDF) (ablation) | baseline + learnable upsampling · Skeleton extraction, ICDAR-2013 synthesized | F-measure: 0.592 -> 0.597 | Helped |
| add MDF (ablation) | baseline + learnable upsampling · Skeleton extraction, ICDAR-2013 synthesized | AMD: 1.44 -> 1.46 | Hurt |
| add regressive dense upsampling convolution (rDUC) (ablation) | baseline + upsampling + MDF · Skeleton extraction, ICDAR-2013 synthesized | F-measure: 0.597 -> 0.610 | Helped |
| add rDUC (ablation) | baseline + upsampling + MDF · Skeleton extraction, ICDAR-2013 synthesized | AMD: 1.46 -> 1.29 | Helped |
| initialize from pre-trained HCCR-CNN9Layer (ablation) | baseline FCN · Skeleton extraction, ICDAR-2013 synthesized | training loss | Helped |
| use FCN skeletons instead of ZhangSuen skeletons for recognition | HCCR-CNN9Layer · ICDAR-2013 offline | accuracy: 69.94% -> 95.53% | Helped |
| add skeleton channel to raw image input (ablation) | HCCR-CNN9Layer · ICDAR-2013 offline | accuracy: 96.81% -> 96.90% | Helped |
| use skeleton input instead of raw image | HCCR-CNN9Layer · ICDAR-2013 offline | accuracy: 96.81% -> 95.53% | Hurt |
| use skeleton input instead of raw when trained on real, tested on synthesized | HCCR-CNN9Layer · ON2OFF-Comp2013 | accuracy: 89.27% -> 94.42% | Helped |
Closest papers by meaning
- Multi-Scale Bidirectional FCN for Object Skeleton Extraction (AAAI 2018 · similarity 0.84)
- Stroke Extraction of Chinese Character Based on Deep Structure Deformable Image Registration (AAAI 2023 · similarity 0.84)
- GAN-Based Unpaired Chinese Character Image Translation via Skeleton Transformation and Stroke Rendering (AAAI 2020 · similarity 0.84)
- Dynamic Semantic-Based Spatial Graph Convolution Network for Skeleton-Based Human Action Recognition (AAAI 2024 · similarity 0.81)
- Revealing Key Details to See Differences: A Novel Prototypical Perspective for Skeleton-based Action Recognition (CVPR 2025 · similarity 0.80)
- USDRL: Unified Skeleton-Based Dense Representation Learning with Multi-Grained Feature Decorrelation (AAAI 2025 · similarity 0.79)
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.