Deep and Fast Approximate Order Independent Transparency
This work addresses a specific challenge in computer graphics for rendering transparent objects, offering a portable and efficient solution, though it is incremental as it builds on existing OIT techniques.
The paper tackles the problem of efficiently computing order independent transparency (OIT) in graphics rendering by introducing a machine learning approach that predicts pixel colors using a pre-trained neural network, resulting in a method that is fast, uses constant memory, and is more accurate than previous approximate methods.
We present a machine learning approach for efficiently computing order independent transparency (OIT). Our method is fast, requires a small constant amount of memory (depends only on the screen resolution and not on the number of triangles or transparent layers), is more accurate as compared to previous approximate methods, works for every scene without setup and is portable to all platforms running even with commodity GPUs. Our method requires a rendering pass to extract all features that are subsequently used to predict the overall OIT pixel color with a pre-trained neural network. We provide a comparative experimental evaluation and shader source code of all methods for reproduction of the experiments.