Semantic scene completion (SSC) aims to infer both the 3D geometry and semantics of a scene from single images. In contrast to prior work on SSC that heavily relies on expensive ground-truth annotations, we approach SSC in an unsupervised setting. Our novel method, SceneDINO, adapts techniques from self-supervised representation learning and 2D unsupervised scene understanding to SSC. Our training exclusively utilizes multi-view consistency self-supervision without any form of semantic or geometric ground truth. Given a single input image, SceneDINO infers the 3D geometry and expressive 3D DINO features in a feed-forward manner. Through a novel 3D feature distillation approach, we obtain unsupervised 3D semantics. In both 3D and 2D unsupervised scene understanding, SceneDINO reaches state-of-the-art segmentation accuracy. Linear probing our 3D features matches the segmentation accuracy of a current supervised SSC approach. Additionally, we showcase the domain generalization and multi-view consistency of SceneDINO, taking the first steps towards a strong foundation for single image 3D scene understanding.
@article{arxiv.2507.06230,
title = {Feed-Forward SceneDINO for Unsupervised Semantic Scene Completion},
author = {Aleksandar Jevtić and Christoph Reich and Felix Wimbauer and Oliver Hahn and Christian Rupprecht and Stefan Roth and Daniel Cremers},
journal= {arXiv preprint arXiv:2507.06230},
year = {2025}
}
Comments
ICCV 2025. Christoph Reich and Aleksandar Jevti\'c - both authors contributed equally. Code: https://github.com/tum-vision/scenedino Project page: https://visinf.github.io/scenedino