Research map / Label-efficient and robust learning / Semi-supervised learning
ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text Generation
Sharon Fogel, Hadar Averbuch-Elor, Sarel Cohen, Shai Mazor, Roee Litman
This paper tried a common change: Add batch / layer normalization
Add conditional batch normalization to the generator · RIMES · WER: 12.46 \pm 0.30 -> 11.68\pm 0.29 Improved
What 431 accepted papers report for this change
- 313 improved
- 39 no clear effect
- 13 mixed
- 66 hurt
- MemoryLLM: Plug-n-Play Interpretable Feed-Forward Memory for Transformers ICML 2026 · use independent layer norm per FFN · training convergence · convergenceImproved
- Angel or Demon: Investigating the Plasticity Interventions' Impact on Backdoor Threats in Deep Reinforcement Learning ICML 2026 · apply Layer Normalization in TM-Scratch · Acrobot and MountainCar · BTP: reduces BTP by over 30%Hurt
- MEAL: A Benchmark for Continual Multi-Agent Reinforcement Learning ICML 2026 · add layer normalization · MEAL Level 1 · soup deliveredImproved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 100k ScrabbleGAN samples and fine-tune on original data | HTR framework · RIMES | WER: 12.29\pm 0.15 -> 11.32\pm 0.31 | Improved |
| Add 100k ScrabbleGAN samples and fine-tune on original data | HTR framework · IAM | WER: 25.10\pm 0.49 -> 23.61\pm 0.36 | Improved |
| Generate 100k samples with CVL style and IAM lexicon | HTR trained on IAM · CVL test | WER: 39.95\pm 0.91 -> 40.24\pm 0.51 | No clear effect |
| Use the CVL lexicon instead of the IAM lexicon in generated training samples | HTR trained on IAM · CVL test | WER: 39.95\pm 0.91 -> 35.98\pm 0.38 | Improved |
| Generate samples with both CVL style and CVL lexicon | HTR trained on IAM · CVL test | WER: 35.98\pm 0.38 -> 29.75\pm 0.67 | Improved |
| Generate samples with both CVL style and CVL lexicon | HTR trained on IAM · CVL test | NED: 14.52\pm 0.51 vs 15.62\pm 0.15 | Improved |
| Use the proposed gradient balancing with alpha 1 (ablation) | ScrabbleGAN-trained HTR · RIMES | WER: 12.64\pm 0.20 -> 11.68\pm 0.29 | Improved |
| Add gradient balancing rather than using no gradient balancing (ablation) | ScrabbleGAN-trained HTR · RIMES | WER: 12.64\pm 0.20 -> 12.83\pm 0.28 | Hurt |
| Set proposed gradient-balancing alpha to 0.1 instead of 1 (ablation) | ScrabbleGAN-trained HTR · RIMES | WER: 11.68\pm 0.29 -> 12.28\pm 0.49 | Hurt |
| Set proposed gradient-balancing alpha to 10 instead of 1 (ablation) | ScrabbleGAN-trained HTR · RIMES | WER: 11.68\pm 0.29 -> 12.03\pm 0.27 | Hurt |
| Remove the LSTM from the CNN recognizer (ablation) | ScrabbleGAN · RIMES | WER: 13.80\pm 0.30 -> 11.68\pm 0.29 | Improved |
| Use a CNN recognizer without a recurrent head instead of a CRNN recognizer (ablation) | ScrabbleGAN · RIMES | WER: 12.18\pm 0.24 -> 11.68\pm 0.29 | Improved |
Closest papers by meaning
- Handwritten Text Recognition: A Survey (TPAMI 2025 · similarity 0.87)
- MetaWriter: Personalized Handwritten Text Recognition Using Meta-Learned Prompt Tuning (CVPR 2025 · similarity 0.87)
- On the Generalization of Handwritten Text Recognition Models (CVPR 2025 · similarity 0.86)
- SemiETS: Integrating Spatial and Content Consistencies for Semi-Supervised End-to-end Text Spotting (CVPR 2025 · similarity 0.85)
- Zero-Shot Styled Text Image Generation, but Make It Autoregressive (CVPR 2025 · similarity 0.85)
- Learning Textual Prompts for Open-World Semi-Supervised Learning (CVPR 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.
