Research map / Detection and segmentation / Open-set and few-shot detection
Generative Shape Models: Joint Text Recognition and Segmentation with Very Little Training Data
Xinghua Lou, Ken Kansky, Wolfgang Lehrach, CC Laan, Bhaskara Marthi, D. Phoenix, Dileep George
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 |
|---|---|---|---|
| replace PhotoOCR with generative shape model | Word parsing · ICDAR 2013 | accuracy: 84.3% -> 86.2% | Improved |
| replace PhotoOCR with generative shape model | Word parsing · SVT | accuracy: 78.0% -> 80.7 | Improved |
| use 1,406 training images instead of millions | Generative shape model · ICDAR 2013 | training data size: 7,900,000 -> 1,406 | Improved |
| replace best reported character detector with generative shape model | Character detection · ICDAR 2013 Segmentation | recall: 68.7% -> 95.0% | Improved |
| greedy font selection instead of training on all fonts | Shape model training · Google Fonts | training images: 25584 -> 776 | Improved |
| turn off the language model (ablation) | Generative shape model · ICDAR 2013 | performance: approximately 15% performance drop | Hurt |
| turn off the language model (ablation) | PhotoOCR · ICDAR 2013 | performance: more than 40% performance drop | Hurt |
| vary scale of test images | Shape model · Training image transforms | robust fitting: 130% and 70% | No clear effect |
| rotate test images | Shape model · Training image transforms | robust fitting: -20 and +20 degrees | No clear effect |
Closest papers by meaning
- Learning Shape-Aware Embedding for Scene Text Detection (CVPR 2019 · similarity 0.88)
- Towards Unconstrained End-to-End Text Spotting (ICCV 2019 · similarity 0.87)
- Boosting Semi-Supervised Scene Text Recognition via Viewing and Summarizing (NeurIPS 2024 · similarity 0.85)
- Intriguing Properties of Generative Classifiers (ICLR 2024 · similarity 0.85)
- LRANet: Towards Accurate and Efficient Scene Text Detection with Low-Rank Approximation Network (AAAI 2024 · similarity 0.85)
- A Generative Approach at the Instance-Level for Image Segmentation Under Limited Training Data Conditions (Student Abstract) (AAAI 2025 · 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.
