DeepSurfels: Learning Online Appearance Fusion
This work addresses the problem of efficiently representing and updating high-frequency textures in 3D scenes for computer graphics and vision researchers.
This paper introduces DeepSurfels, a hybrid scene representation that combines explicit and neural components to encode geometry and appearance. It enables online appearance fusion from RGB images, outperforming classical and recent learning-based methods in runtime, generalization, and scalability.
We present DeepSurfels, a novel hybrid scene representation for geometry and appearance information. DeepSurfels combines explicit and neural building blocks to jointly encode geometry and appearance information. In contrast to established representations, DeepSurfels better represents high-frequency textures, is well-suited for online updates of appearance information, and can be easily combined with machine learning methods. We further present an end-to-end trainable online appearance fusion pipeline that fuses information from RGB images into the proposed scene representation and is trained using self-supervision imposed by the reprojection error with respect to the input images. Our method compares favorably to classical texture mapping approaches as well as recent learning-based techniques. Moreover, we demonstrate lower runtime, im-proved generalization capabilities, and better scalability to larger scenes compared to existing methods.