Research map / Video / Video segmentation and tracking
MT-VAE: Learning Motion Transformations to Generate Multimodal Human Dynamics
Xinchen Yan, Akash Rastogi, Ruben Villegas, Kalyan Sunkavalli, Eli Shechtman, Sunil Hadap, Ersin Yumer, Honglak Lee
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 |
|---|---|---|---|
| use additive transformation instead of vector concatenation | MT-VAE · Aff-Wild (facial expression) | R-MSE test: 0.73 -> 0.47 | Improved |
| use additive transformation instead of vector concatenation | MT-VAE · Aff-Wild (facial expression) | Test CLL: 2.817 -> 3.147 | Improved |
| use additive transformation instead of vector concatenation | MT-VAE · Human3.6M (64-step 2D joints) | S-MSE test: 2.84 -> 2.87 | No clear effect |
| add motion coherence objective (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | R-MSE test: 1.01 -> 0.75 | Improved |
| add motion coherence objective (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | Test CLL: 1.012 -> 1.141 | Improved |
| add cycle consistency objective (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | R-MSE test: 1.18 -> 0.75 | Improved |
| add cycle consistency objective (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | S-MSE test: 2.71 -> 2.87 | Hurt |
| add cycle consistency objective (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | Test CLL: 0.927 -> 1.141 | Improved |
| condition transformation vector on input feature (vs context-free) (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | S-MSE test: 4.05 -> 2.87 | Improved |
| condition transformation vector on input feature (vs context-free) (ablation) | MT-VAE (add) · Human3.6M (64-step 2D joints) | R-MSE test: 0.31 -> 0.75 | Hurt |
| add latent variable model (MT-VAE add) over deterministic Prediction LSTM | Prediction LSTM · Human3.6M (64-step 2D joints) | Test CLL: 0.602 -> 1.141 | Improved |
| use MT-VAE (add) instead of Vanilla VAE | Vanilla VAE · Human3.6M (64-step 2D joints) | S-MSE test: 3.88 -> 2.87 | Improved |
Closest papers by meaning
- Action-Conditioned 3D Human Motion Synthesis With Transformer VAE (ICCV 2021 · similarity 0.91)
- A Unified 3D Human Motion Synthesis Model via Conditional Variational Auto-Encoder (ICCV 2021 · similarity 0.90)
- Large Motion Model for Unified Multi-Modal Motion Generation (ECCV 2024 · similarity 0.89)
- Generating Human Motion From Textual Descriptions With Discrete Representations (CVPR 2023 · similarity 0.89)
- Shape My Moves: Text-Driven Shape-Aware Synthesis of Human Motions (CVPR 2025 · similarity 0.88)
- AMD: Autoregressive Motion Diffusion (AAAI 2024 · similarity 0.87)
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.
