Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation 事件

Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation arXiv:2605.25220v1 Announce Type: new Abstract: High-fidelity 3D Gaussian head avatar generation is critical for applications such as AR/VR, telepresence, and digital humans. Existing methods depend on multi-view datasets, 3D captures, or intermediate 2D view synthesis. In contrast, we learn both conditional and unconditional 3D head models from randomly sampled 2D images alone, without using multi-view data, 3D supe