Research map / Video / Video generation
GReg: Geometry-Aware Region Refinement for Sign Language Video Generation
Tongkai Shi, Lianyu Hu, Fanhua Shang, Liqing Gao, Wei Feng
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 regional prior generation (ablation) | GReg · HamNoSys | FVD: 458.3 -> 391.3 | Improved |
| add gradient-enhanced refinement (ablation) | GReg · HamNoSys | FVD: 441.4 -> 391.3 | Improved |
| add adversarial training on regions and gradients (ablation) | GReg · HamNoSys | FVD: 434.6 -> 391.3 | Improved |
| use gradient (not coordinates) in adversarial training (ablation) | GReg · HamNoSys | FVD: 423.8 -> 391.3 | Improved |
| gradient-based prior constraint instead of grayscale (ablation) | GReg RPG · HamNoSys | FVD: 411.3 -> 391.3 | Improved |
| gradient-based prior constraint instead of RGB (ablation) | GReg RPG · HamNoSys | FVD: 410.3 -> 391.3 | Improved |
| gradient cross-attention query instead of adding gradient feature (ablation) | GReg GR · HamNoSys | FVD: 411.5 -> 391.3 | Improved |
| gradient cross-attention query instead of self-attention (ablation) | GReg GR · HamNoSys | FVD: 401.3 -> 391.3 | Improved |
| SMPL-X normal maps instead of skeleton (ablation) | GReg · HamNoSys | FVD: 422.1 -> 391.3 | Improved |
| SMPL-X normal maps instead of heatmap (ablation) | GReg · HamNoSys | FVD: 412.3 -> 391.3 | Improved |
| SMPL-X normal maps instead of skeleton (ablation) | GReg · HamNoSys | AKD: 0.02403 -> 0.02429 | Hurt |
| use convolutional layers in regional experts instead of modulated style design (ablation) | GReg RPG · HamNoSys | FVD: 401.3 -> 391.3 | Hurt |
Closest papers by meaning
- Advanced Sign Language Video Generation with Compressed and Quantized Multi-Condition Tokenization (NeurIPS 2025 · similarity 0.88)
- Geo-Sign: Hyperbolic Contrastive Regularisation for Geometrically Aware Sign Language Translation (NeurIPS 2025 · similarity 0.87)
- Lost in Translation, Found in Embeddings: Sign Language Translation and Alignment (NeurIPS 2026 · similarity 0.86)
- Sign-IDD: Iconicity Disentangled Diffusion for Sign Language Production (AAAI 2025 · similarity 0.86)
- Conditional Variational Autoencoder for Sign Language Translation with Cross-Modal Alignment (AAAI 2024 · similarity 0.86)
- Discrete to Continuous: Generating Smooth Transition Poses from Sign Language Observations (CVPR 2025 · similarity 0.86)
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.
