4-D Epanechnikov Mixture Regression in Light Field Image Compression

This project is a classic academic research artifact associated with a 2021 arXiv paper. While the mathematical approach of using a 4-D Epanechnikov kernel for light field compression is a novel niche application, the project shows zero signs of commercial or open-source traction (0 stars, 5 forks, stagnant for ~4.6 years). In the competitive landscape of light field and 3D representation, the industry has largely pivoted away from traditional statistical mixture models toward Neural Radiance Fields (NeRFs), 3D Gaussian Splatting, and deep neural compression (e.g., JPEG Pleno Part 2). The defensibility is nearly non-existent as the code serves only as a proof-of-concept for a specific paper. Frontier labs have no interest in this specific kernel-based regression method, as they are solving compression via generative modeling. The primary threat is not platform domination, but technological obsolescence; current AI-driven compression techniques significantly outperform these older statistical regression methods in both reconstruction quality and bitrate efficiency. Any potential value is locked in the mathematical insights rather than the software implementation.