Research map / Video / Video generation
Interactive Conversational Head Generation
Mohan Zhou, Yalong Bai, Wei Zhang, Ting Yao, Tiejun Zhao
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 |
|---|---|---|---|
| use responsive listener conditioned on speaker audio and face instead of random motion | Random listener · ViCo responsive listening | ExpFD: 16.435 -> 15.149 | Improved |
| use responsive listener instead of random motion | Random listener · ViCo-X responsive listening | ExpFD: 13.739 -> 10.656 | Improved |
| use responsive listener instead of mirroring the speaker | Mirror listener · ViCo-X responsive listening | ExpFD: 17.635 -> 10.656 | Improved |
| use responsive listener instead of mirroring the speaker | Mirror listener · ViCo responsive listening | AngleFD: 10.155 -> 7.963 | Improved |
| add listener visual signals to audio-driven speaker (ablation) | Audio-only speaker · ViCo expressive talking head | AVOffset: 0.376 -> 0.094 | Improved |
| add listener visual signals to audio-driven speaker (ablation) | Audio-only speaker · ViCo-X expressive talking head | AVOffset: 0.636 -> 0.083 | Improved |
| add listener visual signals to audio-driven speaker (ablation) | Audio-only speaker · ViCo expressive talking head | ExpFD: 17.486 -> 16.604 | Improved |
| add listener visual signals to audio-driven speaker (ablation) | Audio-only speaker · ViCo expressive talking head | LMD: 13.505 -> 14.366 | Hurt |
| add listener visual signals to audio-driven speaker (ablation) | Audio-only speaker · ViCo-X expressive talking head | AVConf: 2.524 -> 2.755 | Improved |
| add listener visual signals to audio-driven speaker (ablation) | Audio-only speaker · ViCo expressive talking head | FID: 45.581 -> 46.181 | No clear effect |
| use role-switching conversational agent instead of separate best listener plus speaker | Listener + Speaker baseline · ViCo-X conversational head generation | ExpFD: 13.954 -> 13.293 | Improved |
| use role-switching conversational agent instead of separate listener plus speaker | Listener + Speaker baseline · ViCo-X conversational head generation | AVOffset: 0.193 -> 0.362 | Hurt |
Closest papers by meaning
- DualTalk: Dual-Speaker Interaction for 3D Talking Head Conversations (CVPR 2025 · similarity 0.90)
- INFP: Audio-Driven Interactive Head Generation in Dyadic Conversations (CVPR 2025 · similarity 0.88)
- CustomListener: Text-guided Responsive Interaction for User-friendly Listening Head Generation (CVPR 2024 · similarity 0.86)
- EmotiveTalk: Expressive Talking Head Generation through Audio Information Decoupling and Emotional Video Diffusion (CVPR 2025 · similarity 0.85)
- Talking Turns: Benchmarking Audio Foundation Models on Turn-Taking Dynamics (ICLR 2025 · similarity 0.85)
- CoVoMix: Advancing Zero-Shot Speech Generation for Human-like Multi-talker Conversations (NeurIPS 2024 · similarity 0.85)
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.
