pix2gestalt: Amodal Segmentation by Synthesizing Wholes
This addresses the challenge of object perception under occlusion for computer vision applications, offering a zero-shot solution that enhances existing methods.
The paper tackles the problem of zero-shot amodal segmentation, where objects are partially occluded, by introducing pix2gestalt, a framework that estimates whole object shapes and appearances using a conditional diffusion model trained on synthetic data. The approach outperforms supervised baselines on benchmarks and improves object recognition and 3D reconstruction methods in occluded scenarios.
We introduce pix2gestalt, a framework for zero-shot amodal segmentation, which learns to estimate the shape and appearance of whole objects that are only partially visible behind occlusions. By capitalizing on large-scale diffusion models and transferring their representations to this task, we learn a conditional diffusion model for reconstructing whole objects in challenging zero-shot cases, including examples that break natural and physical priors, such as art. As training data, we use a synthetically curated dataset containing occluded objects paired with their whole counterparts. Experiments show that our approach outperforms supervised baselines on established benchmarks. Our model can furthermore be used to significantly improve the performance of existing object recognition and 3D reconstruction methods in the presence of occlusions.