Research map / 3D vision / 3D generation
Making Avatars Interact: Towards Text-Driven Human-Object Interaction for Controllable Talking Avatars
youliang zhang, zhengguang zhou, Zhentao Yu, Ziyao Huang, Teng Hu, Sen Liang, Guozhen Zhang, Ziqiao Peng, Shunkai Li, Yi Chen, Zixiang Zhou, Yuan Zhou and 2 more
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 PIM motion planning (joint generation) (ablation) | AIM-only video generator · GroundedInter | VLM-QA: 26.19 -> 28.89 | Improved |
| add PIM motion planning (joint generation) (ablation) | AIM-only video generator · GroundedInter | Hand Quality (HQ): 0.711 -> 0.925 | Improved |
| use joint training instead of cascaded frozen PIM (ablation) | InteractAvatar · GroundedInter | VLM-QA: 28.19 -> 28.89 | Improved |
| use layer-wise injection instead of last-layer-only injection (ablation) | InteractAvatar · GroundedInter | VLM-QA: 28.31 -> 28.89 | Improved |
| use zero-initialized residual injection instead of simple addition (ablation) | InteractAvatar · GroundedInter | VLM-QA: 28.46 -> 28.89 | Improved |
| add modified RoPE for ref image (ablation) | PIM · GroundedInter | VLM-QA: 26.36 -> 27.15 | Improved |
| add detection data in PIM pre-training (ablation) | PIM · GroundedInter | VLM-QA: 27.15 -> 27.87 | Improved |
| add Det.+Cont. data in PIM pre-training (ablation) | PIM · GroundedInter | VLM-QA: 27.87 -> 28.89 | Improved |
| prepend instead of append ref image (ablation) | PIM · GroundedInter | VLM-QA: 26.71 -> 27.15 | Improved |
| use RGB motion instead of 2D (x,y) motion (ablation) | InteractAvatar · GroundedInter | VLM-QA: 26.07 -> 28.89 | Improved |
| add audio-only pre-training stage before joint audio-motion training (ablation) | AIM · GroundedInter | Sync_conf: 4.23 -> 5.49 | Improved |
| train audio first then add motion instead of motion first (ablation) | AIM · GroundedInter | Sync_conf: 3.98 -> 5.49 | Improved |
Closest papers by meaning
- GAIA: Zero-shot Talking Avatar Generation (ICLR 2024 · similarity 0.89)
- AvatarGO: Zero-shot 4D Human-Object Interaction Generation and Animation (ICLR 2025 · similarity 0.89)
- InstructAvatar: Text-Guided Emotion and Motion Control for Avatar Generation (AAAI 2025 · similarity 0.89)
- Let Them Talk: Audio-Driven Multi-Person Conversational Video Generation (NeurIPS 2025 · similarity 0.87)
- DAWN: Dynamic Frame Avatar with Non-autoregressive Diffusion Framework for Talking head Video Generation (ICLR 2025 · similarity 0.87)
- VLOGGER: Multimodal Diffusion for Embodied Avatar Synthesis (CVPR 2025 · 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.
