Research map / 3D vision / 3D generation
LOOM: Weaving Geometry-Consistent Human-Object Interaction Videos via Progressive Curriculum Learning
Bangya Liu, Zelin Zhao, Ziyang Song, Suman Banerjee, Xinyu Gong
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 RCM cache and per-frame RCM conditioning (ablation) | ByteLOOM · Mani4D-Test | Obj-IoU: 0.6619 -> 0.8288 | Improved |
| add RCM cache (ablation) | ByteLOOM · Mani4D-Test | MEt3R (lower is better): 0.0540 -> 0.0454 | Improved |
| add RCM cache (ablation) | ByteLOOM · Mani4D-Test | T-SSIM: 0.9063 -> 0.9147 | Improved |
| add RCM cache (ablation) | ByteLOOM · Mani4D-Test | LMD (lower is better): 0.1878 -> 0.1427 | Improved |
| add hand-object curriculum II (ablation) | ByteLOOM I + III · Mani4D-Test | Obj-IoU: 0.7627 -> 0.8288 | Improved |
| add hand-object curriculum II (ablation) | ByteLOOM I + III · Mani4D-Test | LMD (lower is better): 0.2054 -> 0.1427 | Improved |
| add hand-object curriculum II (ablation) | ByteLOOM I + III · Mani4D-Test | Face-Cos: 0.8648 -> 0.8891 | No clear effect |
| add optional object NVS curriculum (ablation) | ByteLOOM I + II + III · Mani4D-Test | MEt3R (lower is better): 0.0454 -> 0.0429 | Improved |
| add optional object NVS curriculum (ablation) | ByteLOOM I + II + III · Mani4D-Test | LMD (lower is better): 0.1427 -> 0.1861 | Hurt |
| add optional object NVS curriculum (ablation) | ByteLOOM I + II + III · Mani4D-Test | Obj-IoU: 0.8288 -> 0.7770 | Hurt |
| use ByteLOOM instead of AnchorCrafter with novel human references | ByteLOOM · Mani4D-Test with novel human references | Obj-IoU: 0.2435 -> 0.8212 | Improved |
Closest papers by meaning
- InterAct: Advancing Large-Scale Versatile 3D Human-Object Interaction Generation (CVPR 2025 · similarity 0.90)
- SViMo: Synchronized Diffusion for Video and Motion Generation in Hand-object Interaction Scenarios (NeurIPS 2025 · similarity 0.90)
- Real2Sim in HOI: Toward Physically Plausible HOI Reconstruction from Monocular Videos (NeurIPS 2026 · similarity 0.89)
- Template Free Reconstruction of Human-object Interaction with Procedural Interaction Generation (CVPR 2024 · similarity 0.88)
- InteractAnything: Zero-shot Human Object Interaction Synthesis via LLM Feedback and Object Affordance Parsing (CVPR 2025 · similarity 0.88)
- InterMimic: Towards Universal Whole-Body Control for Physics-Based Human-Object Interactions (CVPR 2025 · 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.
