Event
Kuaishou Kling AI and universities open-source MetaView single-image novel view synthesis framework
1 reports · 1 sources
The Kuaishou Kling AI team, together with Nanyang Technological University and HKUST (Guangzhou), proposed MetaView, a diffusion model-based single-image novel view synthesis framework, and has open-sourced the code, model weights, and an online demo. The method is based on the MM-DiT architecture, injecting implicit geometric priors extracted by a feed-forward geometry-aware network through parallel attention streams, and uses metric depth to anchor scale-aware RoPE, to mitigate geometric collapse and scale drift under large viewpoint changes. According to the team, on three datasets—DL3DV, RealEstate10K, and Sekai-Real-Walk-HQ—MetaView was compared with 7 baseline models including ViewCrafter, GEN3C, and Voyager, and achieved SOTA on 11 of 12 core evaluation dimensions.
Report timeline
- Kuaishou Kling AI and NTU Open-Source MetaView, a Single-Image Novel View Synthesis Framework, Reaching SOTA on 11 of 12 Metrics
The Kuaishou Kling AI team, together with Nanyang Technological University and HKUST (Guangzhou), proposed MetaView, a diffusion-based single-image novel view synthesis framework, and has open-sourced the code, model weights, and an online demo. The method is built on the MM-DiT architecture, injects implicit geometric priors extracted by a feed-forward geometry-aware network through parallel attention streams, and uses metric depth to anchor scale-aware RoPE, mitigating geometric collapse and scale drift under large viewpoint changes.
CVer · 微信公众号 ·