Research map / 3D vision / Human pose estimation
Structured Prediction Helps 3D Human Motion Modelling
Emre Aksan, Manuel Kaufmann, Otmar Hilliges
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 SP-layer | RNN · AMASS 400 ms | Euler angle error: 16.29 -> 12.84 | Improved |
| add SP-layer | QuaterNet · AMASS 400 ms | Euler angle error: 14.54 -> 13.43 | Improved |
| add SP-layer | Seq2seq-dropout · AMASS 400 ms | Euler angle error: 16.13 -> 14.23 | Improved |
| add SP-layer | Seq2seq-sampling · AMASS 400 ms | Euler angle error: 17.33 -> 14.86 | Improved |
| add SP-layer | Seq2seq · AMASS 400 ms | Euler angle error: 17.84 -> 16.43 | Improved |
| add SP-layer | RNN · AMASS 400 ms | positional error: 1.43 -> 0.96 | Improved |
| add SP-layer | QuaterNet · H3.6M walking 400 ms | Euler angle error: 0.62 -> 0.61 | No clear effect |
| add SP-layer | Seq2seq-sampling-sup · H3.6M walking 400 ms | Euler angle error: 0.81 -> 0.61 | Improved |
| add SP-layer | RNN · H3.6M walking 400 ms | Euler angle error: 0.89 -> 0.78 | Improved |
| add per-joint loss without SP-layer (ablation) | RNN · AMASS 400 ms | Euler angle error: 16.44 -> 13.13 | Improved |
| add per-joint loss without SP-layer (ablation) | RNN · H3.6M walking 400 ms | Euler angle error: 0.900 -> 0.950 | Hurt |
| add per-joint loss without SP-layer (ablation) | Seq2seq · AMASS 400 ms | Euler angle error: 17.84 -> 19.78 | Hurt |
Closest papers by meaning
- Human Motion Prediction via Learning Local Structure Representations and Temporal Dependencies (AAAI 2019 · similarity 0.91)
- Learning Dynamic Relationships for 3D Human Motion Prediction (CVPR 2020 · similarity 0.91)
- Learning Trajectory Dependencies for Human Motion Prediction (ICCV 2019 · similarity 0.90)
- Learning Progressive Joint Propagation for Human Motion Prediction (ECCV 2020 · similarity 0.90)
- Investigating Pose Representations and Motion Contexts Modeling for 3D Motion Prediction (TPAMI 2022 · similarity 0.89)
- SimpliHuMoN: Simplifying Human Motion Prediction (NeurIPS 2026 · 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.
