Research map / Video / Video generation
SignGen: End-to-End Sign Language Video Generation with Latent Diffusion
Fan Qi, Yu Duan, Changsheng Xu, Huaiwen Zhang
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 multi-modal conditions (flow, pose, depth) for sign video (ablation) | SignGen · RWTH-2014 | FVD: 1014 -> 579 | Improved |
| add motion-aware sign video generation architecture (ablation) | SignGen · RWTH-2014 | FVD: 631 -> 579 | Improved |
| add motion conditions progressively (one to two to three) (ablation) | SignGen · RWTH-2014 | FVD: 726 -> 628 -> 579 | Improved |
| add motion conditions progressively (one to three) (ablation) | SignGen · RWTH-2014 (back translation test) | BLEU-4: 12.45 -> 19.71 | Improved |
| remove the anchor image at inference (ablation) | SignGen · qualitative | semantic consistency | Hurt |
| use end-to-end text-to-video rather than cascaded Text2Gloss2Pose pipelines | sign language generation · RWTH-2014 Text2Pose (test) | BLEU-4: 9.68 -> 19.71 | Improved |
Closest papers by meaning
- Advanced Sign Language Video Generation with Compressed and Quantized Multi-Condition Tokenization (NeurIPS 2025 · similarity 0.91)
- Discrete to Continuous: Generating Smooth Transition Poses from Sign Language Observations (CVPR 2025 · similarity 0.89)
- Lost in Translation, Found in Context: Sign Language Translation with Contextual Cues (CVPR 2025 · similarity 0.89)
- Scaling Sign Language Translation (NeurIPS 2024 · similarity 0.89)
- Lost in Translation, Found in Embeddings: Sign Language Translation and Alignment (NeurIPS 2026 · similarity 0.89)
- Bridging Sign and Spoken Languages: Pseudo Gloss Generation for Sign Language Translation (NeurIPS 2025 · similarity 0.89)
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.
