SRGS: Super-Resolution 3D Gaussian Splatting
This work addresses the challenge of achieving high-resolution novel view synthesis from low-resolution inputs in 3D reconstruction, which is incremental as it builds upon existing 3DGS methods.
The paper tackles the problem of low-resolution 3D Gaussian Splatting (3DGS) leading to sparse and texture-deficient primitives, hindering high-resolution novel view synthesis (HRNVS), by proposing SRGS, which optimizes in high-resolution space with sub-pixel constraints and a pre-trained 2D super-resolution model, achieving high rendering quality that outperforms state-of-the-art methods on datasets like Mip-NeRF 360 and Tanks & Temples.
Recently, 3D Gaussian Splatting (3DGS) has gained popularity as a novel explicit 3D representation. This approach relies on the representation power of Gaussian primitives to provide a high-quality rendering. However, primitives optimized at low resolution inevitably exhibit sparsity and texture deficiency, posing a challenge for achieving high-resolution novel view synthesis (HRNVS). To address this problem, we propose Super-Resolution 3D Gaussian Splatting (SRGS) to perform the optimization in a high-resolution (HR) space. The sub-pixel constraint is introduced for the increased viewpoints in HR space, exploiting the sub-pixel cross-view information of the multiple low-resolution (LR) views. The gradient accumulated from more viewpoints will facilitate the densification of primitives. Furthermore, a pre-trained 2D super-resolution model is integrated with the sub-pixel constraint, enabling these dense primitives to learn faithful texture features. In general, our method focuses on densification and texture learning to effectively enhance the representation ability of primitives. Experimentally, our method achieves high rendering quality on HRNVS only with LR inputs, outperforming state-of-the-art methods on challenging datasets such as Mip-NeRF 360 and Tanks & Temples. Related codes will be released upon acceptance.