Mipmap-GS: Let Gaussians Deform with Scale-specific Mipmap for Anti-aliasing Rendering
This addresses aliasing issues in 3D rendering for computer vision applications, offering a plug-in solution to enhance existing models, though it is incremental as it builds upon 3DGS.
The paper tackles the problem of zooming degradation in 3D Gaussian Splatting for novel view synthesis by proposing a method to make Gaussians adaptive to arbitrary scales, resulting in average PSNR improvements of 9.25 dB for zoom-in and 10.40 dB for zoom-out on the NeRF Synthetic dataset.
3D Gaussian Splatting (3DGS) has attracted great attention in novel view synthesis because of its superior rendering efficiency and high fidelity. However, the trained Gaussians suffer from severe zooming degradation due to non-adjustable representation derived from single-scale training. Though some methods attempt to tackle this problem via post-processing techniques such as selective rendering or filtering techniques towards primitives, the scale-specific information is not involved in Gaussians. In this paper, we propose a unified optimization method to make Gaussians adaptive for arbitrary scales by self-adjusting the primitive properties (e.g., color, shape and size) and distribution (e.g., position). Inspired by the mipmap technique, we design pseudo ground-truth for the target scale and propose a scale-consistency guidance loss to inject scale information into 3D Gaussians. Our method is a plug-in module, applicable for any 3DGS models to solve the zoom-in and zoom-out aliasing. Extensive experiments demonstrate the effectiveness of our method. Notably, our method outperforms 3DGS in PSNR by an average of 9.25 dB for zoom-in and 10.40 dB for zoom-out on the NeRF Synthetic dataset.