Research map / 3D vision / Face and shape
Expressive Telepresence via Modular Codec Avatars
Hang Chu, Shugao Ma, Fernando De la Torre, Sanja Fidler, Yaser Sheikh
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 learned blending weights instead of equal weights (ablation) | MCA · VR headset face dataset (test sessions), all test data | RMSE: 7.33 -> 6.67 | Improved |
| add end-to-end training (ablation) | MCA · VR headset face dataset (test sessions), all test data | RMSE: 6.67 -> 6.10 | Improved |
| add soft exemplar latent codes with gaussian noise instead of hard classes (ablation) | MCA · VR headset face dataset (test sessions), all test data | RMSE: 6.10 -> 5.71 | Improved |
| expand latent code dimension (ablation) | MCA · VR headset face dataset (test sessions), all test data | RMSE: 5.71 -> 5.64 | Improved |
| add cross-module skip connections (ablation) | MCA · VR headset face dataset (test sessions), all test data | RMSE: 5.64 -> 5.52 | Improved |
| add temporal convolution in the synthesizer (ablation) | MCA · VR headset face dataset (test sessions), all test data | RMSE: 5.52 -> 5.48 | Improved |
| replace holistic Codec Avatar with modular MCA | Codec Avatar · VR headset face dataset (test sessions), overall | RMSE: 5.81 -> 5.48 | Improved |
| replace holistic Codec Avatar with modular MCA | Codec Avatar · VR headset face dataset (test sessions), overall | MAE: 6.54 -> 6.17 | Improved |
| replace holistic Codec Avatar with modular MCA | Codec Avatar · VR headset face dataset (test sessions), overall | SSIM: 0.953 -> 0.956 | Improved |
| replace holistic Codec Avatar with modular MCA | Codec Avatar · VR headset face dataset (test sessions), all persons | Geo. RMSE: 1.37 -> 1.17 | Mixed |
| replace holistic Codec Avatar with modular MCA | Codec Avatar · VR headset face dataset (test sessions), person4 | SSIM: 0.984 -> 0.984 | No clear effect |
| replace holistic Codec Avatar with modular MCA | Codec Avatar · VR headset face dataset (test sessions), person3 | %-better: 0.3 -> 99.7 | Improved |
Closest papers by meaning
- ESCA: Enabling Seamless Codec Avatar Execution through Algorithm and Hardware Co-Optimization for Virtual Reality (NeurIPS 2025 · similarity 0.87)
- High-Fidelity 3D Head Avatars Reconstruction through Spatially-Varying Expression Conditioned Neural Radiance Field (AAAI 2024 · similarity 0.86)
- LUCAS: Layered Universal Codec Avatars (CVPR 2025 · similarity 0.85)
- Codec Avatar Studio: Paired Human Captures for Complete, Driveable, and Generalizable Avatars (NeurIPS 2024 · similarity 0.85)
- Relightable Gaussian Codec Avatars (CVPR 2024 · similarity 0.84)
- RAM-Avatar: Real-time Photo-Realistic Avatar from Monocular Videos with Full-body Control (CVPR 2024 · 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.
