Novel View Synthesis via Depth-guided Skip Connections
This work provides an incremental improvement for researchers working on novel view synthesis, specifically addressing the trade-off between structural consistency and detail preservation.
This paper addresses the problem of novel view synthesis from a single source image. The authors propose an encoder-decoder architecture that regresses target view pixels and maintains details by coupling decoder-aligned feature maps with skip connections guided by a predicted depth map, successfully preserving texture details without distortions.
We introduce a principled approach for synthesizing new views of a scene given a single source image. Previous methods for novel view synthesis can be divided into image-based rendering methods (e.g. flow prediction) or pixel generation methods. Flow predictions enable the target view to re-use pixels directly, but can easily lead to distorted results. Directly regressing pixels can produce structurally consistent results but generally suffer from the lack of low-level details. In this paper, we utilize an encoder-decoder architecture to regress pixels of a target view. In order to maintain details, we couple the decoder aligned feature maps with skip connections, where the alignment is guided by predicted depth map of the target view. Our experimental results show that our method does not suffer from distortions and successfully preserves texture details with aligned skip connections.