Research map / 3D vision / Face and shape
RealDyadic: Synthesizing Realistic Dyadic 3D Dialogue with Neural Appearance Priors
Lei Zhu, Lijian Lin, Ye Zhu, Xuehan Hou, Jiahao Wu, Yu Li, Yunfei Liu, Jie Chen
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 dual-audio fusion module (ablation) | DiffPoseTalk baseline · Real-Dialog test | LVE: 0.269 -> 0.193 | Improved |
| add dual-audio fusion module (ablation) | DiffPoseTalk baseline · Real-Dialog test | PSNR: 18.97 -> 20.17 | Improved |
| add dual-audio fusion module (ablation) | DiffPoseTalk baseline · Real-Dialog test | FDD: 1.607 -> 1.603 | No clear effect |
| add residual audio connection (ablation) | audio fusion · Real-Dialog test | LVE: 0.193 -> 0.233 | Hurt |
| add residual audio connection (ablation) | audio fusion · Real-Dialog test | MOD: 1.281 -> 1.230 | Improved |
| add speaking indicator (ablation) | audio fusion + residual · Real-Dialog test | SLCC: 0.683 -> 0.698 | Improved |
| add speaking indicator (ablation) | audio fusion + residual · Real-Dialog test | MTM: 4.371 -> 4.526 | Hurt |
| add jaw pose loss (ablation) | previous variant · Real-Dialog test | SLCC: 0.698 -> 0.802 | Improved |
| add second-stage 2D photometric supervision with alternating training (ablation) | stage-1 model · Real-Dialog test | LVE: 0.235 -> 0.174 | Improved |
| add second-stage 2D photometric supervision with alternating training (ablation) | stage-1 model · Real-Dialog test | PSNR: 21.20 -> 23.25 | Improved |
| add second-stage 2D photometric supervision with alternating training (ablation) | stage-1 model · Real-Dialog test | SLCC: 0.802 -> 0.791 | Hurt |
| use RealDyadic instead of Spectre tracking as mesh source | lip keypoint accuracy · manually annotated keypoints | MAELip: 9.68 -> 6.23 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DualTalk — DualTalk: Dual-Speaker Interaction for 3D Talking Head Conversations (CVPR 2025) · better on 9 of 10 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Perceptually Accurate 3D Talking Head Generation: New Definitions, Speech-Mesh Representation, and Evaluation Metrics (CVPR 2025 · similarity 0.90)
- FaceTalk: Audio-Driven Motion Diffusion for Neural Parametric Head Models (CVPR 2024 · similarity 0.89)
- Cafe-Talk: Generating 3D Talking Face Animation with Multimodal Coarse- and Fine-grained Control (ICLR 2025 · similarity 0.88)
- GoHD: Gaze-oriented and Highly Disentangled Portrait Animation with Rhythmic Poses and Realistic Expressions (AAAI 2025 · similarity 0.88)
- Probabilistic Speech-Driven 3D Facial Motion Synthesis: New Benchmarks Methods and Applications (CVPR 2024 · similarity 0.88)
- From Audio to Photoreal Embodiment: Synthesizing Humans in Conversations (CVPR 2024 · similarity 0.88)
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.
