Research map / 3D vision / 3D generation
HMD-NeMo: Online 3D Avatar Motion Generation From Sparse Observations
Sadegh Aliakbarian, Fatemeh Saleh, David Collier, Pashmina Cameron, Darren Cosker
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 |
|---|---|---|---|
| replace learned and fixed mask parameters with temporally adaptable mask tokens (TAMT) (ablation) | HMD-NeMo · AMASS hand tracking | MPJPE / MPJVE: 5.59 -> 2.48 MPJPE; 42.81 -> 31.30 MPJVE | Improved |
| add hands expressed in head space to the input (ablation) | HMD-NeMo · AMASS hand tracking | MPJPE: 4.38 -> 3.21 | Improved |
| add velocity inputs (ablation) | HMD-NeMo · AMASS hand tracking | MPJVE: 39.63 -> 35.27 | Improved |
| include GRU in the spatio-temporal encoder (ablation) | HMD-NeMo · AMASS hand tracking | MPJVE: 46.61 -> 31.30 | Improved |
| include transformer encoder in the spatio-temporal encoder (ablation) | HMD-NeMo · AMASS hand tracking | MPJPE: 5.55 -> 2.48 | Improved |
| add post-hoc optimization of predicted pose to observations (ablation) | HMD-NeMo · AMASS motion controllers | MPJPE: 2.07 -> 1.90 (MPJVE 26.07 -> 24.99) | Improved |
| optimize with non-robust L2 energy in hand tracking (ablation) | HMD-NeMo · AMASS hand tracking | MPJPE / MPJVE: MPJPE 2.48 -> 2.32; MPJVE 31.30 -> 33.51 | Mixed |
| optimize with robust energy in hand tracking (ablation) | HMD-NeMo · AMASS hand tracking | MPJPE / MPJVE: MPJPE 2.48 -> 2.37; MPJVE 31.30 -> 31.33 | Mixed |
| add L_SE(3) loss on world-space joint transforms (ablation) | HMD-NeMo · AMASS | MPJPE / MPJVE | Improved |
| add smoothness loss (ablation) | HMD-NeMo · AMASS | MPJVE | Improved |
| add forecasting auxiliary loss for hands (ablation) | HMD-NeMo · AMASS | MPJPE / MPJVE | Improved |
Closest papers by meaning
- HMD-Poser: On-Device Real-time Human Motion Tracking from Scalable Sparse Observations (CVPR 2024 · similarity 0.88)
- From Sparse Signal to Smooth Motion: Real-Time Motion Generation with Rolling Prediction Models (CVPR 2025 · similarity 0.87)
- Stratified Avatar Generation from Sparse Observations (CVPR 2024 · similarity 0.87)
- AvatarGO: Zero-shot 4D Human-Object Interaction Generation and Animation (ICLR 2025 · similarity 0.87)
- FRAME: Floor-aligned Representation for Avatar Motion from Egocentric Video (CVPR 2025 · similarity 0.86)
- RAM-Avatar: Real-time Photo-Realistic Avatar from Monocular Videos with Full-body Control (CVPR 2024 · 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.
