DR-KFS: A Differentiable Visual Similarity Metric for 3D Shape Reconstruction
This addresses the need for more visually accurate 3D reconstructions in computer vision applications, though it is incremental as it builds on existing reconstruction networks.
The paper tackled the problem of improving 3D shape reconstruction quality by introducing a differentiable visual similarity metric that compares multi-view images rendered from shapes, replacing distortion-based losses. The result showed better structural fidelity and visual quality, validated through objective metrics and a perceptual study.
We introduce a differential visual similarity metric to train deep neural networks for 3D reconstruction, aimed at improving reconstruction quality. The metric compares two 3D shapes by measuring distances between multi-view images differentiably rendered from the shapes. Importantly, the image-space distance is also differentiable and measures visual similarity, rather than pixel-wise distortion. Specifically, the similarity is defined by mean-squared errors over HardNet features computed from probabilistic keypoint maps of the compared images. Our differential visual shape similarity metric can be easily plugged into various 3D reconstruction networks, replacing their distortion-based losses, such as Chamfer or Earth Mover distances, so as to optimize the network weights to produce reconstructions with better structural fidelity and visual quality. We demonstrate this both objectively, using well-known shape metrics for retrieval and classification tasks that are independent from our new metric, and subjectively through a perceptual study.