PixelSynth: Generating a 3D-Consistent Experience from a Single Image
This addresses the challenge of creating 3D-consistent experiences for applications like virtual reality or content creation, though it appears incremental by building on recent advancements in differentiable rendering and 3D reasoning.
The paper tackles the problem of generating immersive 3D scenes from a single image by enabling large-angle view synthesis, showing considerable improvement over existing methods in both simulated and real datasets.
Recent advancements in differentiable rendering and 3D reasoning have driven exciting results in novel view synthesis from a single image. Despite realistic results, methods are limited to relatively small view change. In order to synthesize immersive scenes, models must also be able to extrapolate. We present an approach that fuses 3D reasoning with autoregressive modeling to outpaint large view changes in a 3D-consistent manner, enabling scene synthesis. We demonstrate considerable improvement in single image large-angle view synthesis results compared to a variety of methods and possible variants across simulated and real datasets. In addition, we show increased 3D consistency compared to alternative accumulation methods. Project website: https://crockwell.github.io/pixelsynth/