Research map / Label-efficient and robust learning / Recognition and long-tail
Style Transfer in Text: Exploration and Evaluation
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 |
|---|---|---|---|
| Use adversarially trained content representations with style-specific decoders | Multi-decoder · Paper-news title transfer | transfer strength | Helped |
| Use a shared decoder controlled by style embeddings | Style-embedding · Paper-news title transfer | content preservation | Helped |
| Use multi-decoder instead of style-embedding | Style transfer models · Positive-negative review transfer | transfer strength | Helped |
| Increase encoder (decoder) dimension from 32 to 128 (ablation) | Auto-encoder · Paper-news title transfer | content preservation: 0.9118 -> 0.9837 | Helped |
| Increase encoder (decoder) dimension from 32 to 128 (ablation) | Auto-encoder · Paper-news title transfer | transfer strength: 0.035 -> 0.0135 | Hurt |
| Increase encoder (decoder) dimension from 32 to 128 (ablation) | Multi-decoder · Paper-news title transfer | content preservation: 0.8335 -> 0.8705 | Helped |
| Increase encoder (decoder) dimension from 32 to 128 (ablation) | Multi-decoder · Paper-news title transfer | transfer strength: 0.9915 -> 0.842 | Hurt |
| Increase style embedding dimension from 32 to 128 (ablation) | Style-embedding, encoder dimension 32 · Paper-news title transfer | content preservation: 0.8966 -> 0.9435 | Helped |
| Increase style embedding dimension from 32 to 128 (ablation) | Style-embedding, encoder dimension 32 · Paper-news title transfer | transfer strength: 0.593 -> 0.282 | Hurt |
| Increase style embedding dimension from 32 to 128 (ablation) | Style-embedding, encoder dimension 64 · Paper-news title transfer | content preservation: 0.8957 -> 0.9440 | Helped |
| Increase style embedding dimension from 32 to 128 (ablation) | Style-embedding, encoder dimension 64 · Paper-news title transfer | transfer strength: 0.596 -> 0.2745 | Hurt |
| Increase style embedding dimension from 32 to 128 (ablation) | Style-embedding, encoder dimension 128 · Paper-news title transfer | content preservation: 0.8960 -> 0.9439 | Helped |
Closest papers by meaning
- Revision in Continuous Space: Unsupervised Text Style Transfer without Adversarial Learning (AAAI 2020 · similarity 0.90)
- Non-Parallel Text Style Transfer with Self-Parallel Supervision (ICLR 2022 · similarity 0.90)
- ParaGuide: Guided Diffusion Paraphrasers for Plug-and-Play Textual Style Transfer (AAAI 2024 · similarity 0.88)
- StyleStudio: Text-Driven Style Transfer with Selective Control of Style Elements (CVPR 2025 · similarity 0.86)
- Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfer (CVPR 2024 · similarity 0.85)
- SigStyle: Signature Style Transfer via Personalized Text-to-Image Models (AAAI 2025 · similarity 0.85)
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.