Research map / 3D vision / 3D generation
SceMoS: Scene-Aware 3D Human Motion Synthesis by Planning with Geometry-Grounded Tokens
Anindita Ghosh, Vladislav Golyanik, Taku Komura, Philipp Slusallek, Christian Theobalt, Rishabh Dabral
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 two-stage planner/decoder split instead of single transformer (ablation) | SceMoS · TRUMANS | FID: 0.78 -> 0.31 | Improved |
| use DINOv2 instead of CLIP for BEV features (ablation) | SceMoS · TRUMANS | FID: 0.81 -> 0.31 | Improved |
| add trajectory refinement module (ablation) | SceMoS · TRUMANS | FID: 0.53 -> 0.31 | Improved |
| condition VQ-VAE decoder on local heightmap (ablation) | SceMoS tokenizer · TRUMANS | MPJPE (mm): 25.89 -> 21.88 | Improved |
| replace 2D heightmap with 3D voxel grid (ablation) | SceMoS tokenizer · TRUMANS | mean penetration: 1.83 -> 3.88 | Hurt |
| use 64x64 heightmap instead of 32x32 (ablation) | SceMoS tokenizer · TRUMANS | MPJPE (mm): 21.88 -> 22.56 | Hurt |
| use 16x16 heightmap instead of 32x32 (ablation) | SceMoS tokenizer · TRUMANS | mean penetration: 1.83 -> 2.12 | Hurt |
| fuse heightmap with FiLM instead of concatenation (ablation) | SceMoS tokenizer · TRUMANS | MPJPE (mm): 21.88 -> 27.89 | Hurt |
| fuse heightmap with cross-attention instead of concatenation (ablation) | SceMoS tokenizer · TRUMANS | MPJVE (mm): 10.45 -> 13.24 | Hurt |
| SceMoS vs TRUMANS with far fewer scene params | SceMoS · TRUMANS | FID: 0.34 -> 0.31 | No clear effect |
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.
- Humanise — HUMANISE: Language-conditioned Human Motion Generation in 3D Scenes (NeurIPS 2022) · better on 4 of 4 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
- SceneWeaver: All-in-One 3D Scene Synthesis with an Extensible and Self-Reflective Agent (NeurIPS 2025 · similarity 0.87)
- InstructScene: Instruction-Driven 3D Indoor Scene Synthesis with Semantic Graph Prior (ICLR 2024 · similarity 0.87)
- MetaSpatial: Reinforcing 3D Spatial Reasoning in VLMs for the Metaverse (ICLR 2026 · similarity 0.86)
- InfiniBench: Infinite Benchmarking for Visual Spatial Reasoning with Customizable Scene Complexity (CVPR 2026 · similarity 0.86)
- Video Perception Models for 3D Scene Synthesis (NeurIPS 2025 · similarity 0.86)
- S$^2$-MLLM: Boosting Spatial Reasoning Capability of MLLMs for 3D Visual Grounding with Structural Guidance (CVPR 2026 · similarity 0.86)
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.
