EigenGS Representation: From Eigenspace to Gaussian Image Space
This work addresses the computational bottleneck in Gaussian-based image representation for computer vision applications, offering an incremental improvement over existing methods.
The paper tackles the problem of slow optimization for Gaussian-based image representation by introducing EigenGS, which bridges PCA eigenspace with image-space Gaussian representations to enable instant initialization without per-image optimization. The method achieves superior reconstruction quality while reducing parameter count and training time, making it viable for real-time applications.
Principal Component Analysis (PCA), a classical dimensionality reduction technique, and 2D Gaussian representation, an adaptation of 3D Gaussian Splatting for image representation, offer distinct approaches to modeling visual data. We present EigenGS, a novel method that bridges these paradigms through an efficient transformation pipeline connecting eigenspace and image-space Gaussian representations. Our approach enables instant initialization of Gaussian parameters for new images without requiring per-image optimization from scratch, dramatically accelerating convergence. EigenGS introduces a frequency-aware learning mechanism that encourages Gaussians to adapt to different scales, effectively modeling varied spatial frequencies and preventing artifacts in high-resolution reconstruction. Extensive experiments demonstrate that EigenGS not only achieves superior reconstruction quality compared to direct 2D Gaussian fitting but also reduces necessary parameter count and training time. The results highlight EigenGS's effectiveness and generalization ability across images with varying resolutions and diverse categories, making Gaussian-based image representation both high-quality and viable for real-time applications.