Search papers, labs, and topics across Lattice.
This paper introduces Locality-Aware Density Control (LocoADC), a framework designed to optimize Gaussian capacity allocation in 2D Gaussian Splatting for image representation. By leveraging local continuity of reconstruction errors and local similarity among neighboring Gaussians, LocoADC effectively addresses issues of under-reconstruction and redundancy, leading to improved Gaussian utilization. Experimental results demonstrate a significant enhancement in representation quality, achieving a 2.93 dB PSNR improvement over existing methods on the CLIC dataset with the same Gaussian budget.
Efficiently optimizing Gaussian capacity allocation can yield nearly 3 dB improvements in image representation quality without increasing the computational budget.
2D Gaussian Splatting is an attractive direction for image representation due to its explicit formulation, fast rasterization, and favorable decoding efficiency. The representation quality of this paradigm depends on the proper allocation of Gaussian capacity to the demanding regions. However, existing methods fail to allocate Gaussian capacity efficiently during optimization: under-reconstructed content is often refined in a fragmented pixel-wise manner, while neighboring optimized Gaussians with similar attributes are redundantly retained. This inefficiency motivates the need for a density control framework that jointly addresses insufficient allocation in under-reconstructed regions and redundant allocation in over-reconstructed regions. Our key insight is that this framework should exploit two complementary forms of locality: the local continuity of reconstruction errors in image space for improved Gaussian allocation, and the local similarity of neighboring Gaussians in Gaussian space for redundant elimination. Based on this insight, we propose Locality-Aware Density Control (LocoADC), a plug-and-play framework that improves Gaussian capacity utilization through Region-wise Gaussian Densification (RGD) and Similarity-Driven Gaussian Merging (SDGM) strategies, together with a local color consistency constraint for more reliable merging. Extensive experiments on diverse datasets show that LocoADC consistently improves multiple baselines by enabling more effective local Gaussian allocation, including a 2.93 dB PSNR gain over GI on the CLIC dataset under the same 30k Gaussian budget. Code is available at: \textit{https://github.com/ChenJiaCong-1005/LocoADC}.
