pixelNeRF: Neural Radiance Fields from One or Few Images
Authors:
Preprint
Abstract
We propose pixelNeRF, a learning framework that predicts a continuous neural scene representation conditioned on one or few input images. The existing approach for constructing neural radiance fields involves optimizing the representation to every scene independently, requiring many calibrated views and significant compute time. We take a step towards resolving these shortcomings by introducing an architecture that conditions a NeRF on image inputs in a fully convolutional manner. This allows the network to be trained across multiple scenes to learn a scene prior, enabling it to perform novel view synthesis in a feed-forward manner from a sparse set of views (as few as one). Leveraging the volume rendering approach of NeRF, our model can be trained directly from images with no explicit 3D supervision. We conduct extensive experiments on ShapeNet benchmarks for single image novel view synthesis tasks with held-out objects as well as entire unseen categories. We further demonstrate the flexibility of pixelNeRF by demonstrating it on multi-object ShapeNet scenes and real scenes from the DTU dataset. In all cases, pixelNeRF outperforms current state-of-the-art baselines for novel view synthesis and single image 3D reconstruction. For the video and code, please visit the project website: https://alexyu.net/pixelnerfKeywords
All Related Versions
This publication has 25 references indexed in Scilit:
- DeepSDF: Learning Continuous Signed Distance Functions for Shape Representation2019
- Semantic Image Synthesis With Spatially-Adaptive Normalization2019
- MVSNet: Depth Inference for Unstructured Multi-view Stereo2018
- Pixel2Mesh: Generating 3D Mesh Models from Single RGB Images2018
- The Unreasonable Effectiveness of Deep Features as a Perceptual Metric2018
- Pixels, Voxels, and Views: A Study of Shape Representations for Single View 3D Object Shape Prediction2018
- Neural 3D Mesh Renderer2018
- View Synthesis by Appearance Flow2016
- Pixelwise View Selection for Unstructured Multi-View Stereo2016
- Large Scale Multi-view Stereopsis Evaluation2014