Research map / 3D vision / 3D generation
InterPet4D: A Multimodal 4D Human-Pet Interaction Dataset for Pet Motion Generation
YICHEN PENG, Jyun-Ting Song, Chen-Chieh Liao, Kris Kitani, Hideki Koike, Erwin Wu
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 PetVAE discrete motion tokenization (ablation) | IPMG · InterPet4D val | FID_k: 14.15 -> 11.21 | Improved |
| use MAA instead of causal MHA (ablation) | IPMG · InterPet4D val | FID_k / FID_s: 13.83/15.44 to 11.21/12.96 | Improved |
| use IPMG instead of plain-MHA seq2seq transformer | seq2seq-Transformer · InterPet4D val | FID_k: 21.22 -> 11.21 | Improved |
| use IPMG instead of generic diffusion transformer | DiT · InterPet4D val | FID_k: 64.37 -> 11.21 | Improved |
| add hand input to body input (ablation) | IPMG · InterPet4D val | FID_k: 13.48 -> 11.21 | Improved |
| add body input to hand input (ablation) | IPMG · InterPet4D val | FID_k: 17.92 -> 11.21 | Improved |
| add hand input to body-only model (ablation) | IPMG · InterPet4D val | R-Precision hand Top-3: 0.44 -> 0.63 | Improved |
| add body input to hand-only model (ablation) | IPMG · InterPet4D val | R-Precision body Top-3: 0.36 -> 0.59 | Improved |
| add bone constraint to PetVAE (ablation) | PetVAE · InterPet4D val | FID: 12.53 -> 9.20 | Improved |
| add bone constraint to PetVAE (ablation) | PetVAE · InterPet4D val | MPJPE: 7.26 -> 6.42 | Improved |
| add bone constraint to PetVAE (ablation) | PetVAE · InterPet4D val | velocity error: 1.79 -> 1.03 | Improved |
| add audio conditioning (ablation) | IPMG · InterPet4D val | FID_k: 13.69 -> 11.21 | Improved |
Closest papers by meaning
- Animal-Bench: Benchmarking Multimodal Video Models for Animal-centric Video Understanding (NeurIPS 2024 · similarity 0.84)
- Harmony4D: A Video Dataset for In-The-Wild Close Human Interactions (NeurIPS 2024 · similarity 0.83)
- Web-Scale Collection of Video Data for 4D Animal Reconstruction (NeurIPS 2025 · similarity 0.83)
- HOI-M^3: Capture Multiple Humans and Objects Interaction within Contextual Environment (CVPR 2024 · similarity 0.82)
- InterAct: Advancing Large-Scale Versatile 3D Human-Object Interaction Generation (CVPR 2025 · similarity 0.82)
- InterDreamer: Zero-Shot Text to 3D Dynamic Human-Object Interaction (NeurIPS 2024 · similarity 0.82)
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.
