Neural Directional Encoding for Efficient and Accurate View-Dependent Appearance Modeling
About
Novel-view synthesis of specular objects like shiny metals or glossy paints remains a significant challenge. Not only the glossy appearance but also global illumination effects, including reflections of other objects in the environment, are critical components to faithfully reproduce a scene. In this paper, we present Neural Directional Encoding (NDE), a view-dependent appearance encoding of neural radiance fields (NeRF) for rendering specular objects. NDE transfers the concept of feature-grid-based spatial encoding to the angular domain, significantly improving the ability to model high-frequency angular signals. In contrast to previous methods that use encoding functions with only angular input, we additionally cone-trace spatial features to obtain a spatially varying directional encoding, which addresses the challenging interreflection effects. Extensive experiments on both synthetic and real datasets show that a NeRF model with NDE (1) outperforms the state of the art on view synthesis of specular objects, and (2) works with small networks to allow fast (real-time) inference. The project webpage and source code are available at: \url{https://lwwu2.github.io/nde/}.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Novel View Synthesis | Synthetic dynamic scenes | PSNR37.19 | 19 | |
| Novel View Synthesis | Teaser scene Synthetic | PSNR34.08 | 5 | |
| View Synthesis | Ref-NeRF Synthetic Scenes (test) | FLIP (Mat.)0.039 | 5 | |
| View Synthesis | Shiny Blender and NeRF Synthetic Materials scene | PSNR (Material)31.53 | 4 | |
| Novel View Synthesis | Ref-NeRF real dataset Garden Spheres scene | PSNR23.63 | 3 |