Smoothness, Synthesis, and Sampling: Re-thinking Unsupervised Multi-View Stereo with DIV Loss
Alex Rich, Noah Stier, Pradeep Sen, Tobias Höllerer
ECCV 2024 (Oral Presentation)
project page  /  PDF  /  code

We introduce DIV Loss, a novel loss formulation for unsupservised multi-view stereo. The key advantage of our DIV loss is that it can be easily dropped into existing unsupervised MVS training pipelines, resulting in significant quantitative and qualitative improvements.

3DVNet: Multi-View Depth Prediction and Volumetric Refinement
Alexander Rich, Noah Stier, Pradeep Sen, Tobias Höllerer
3DV 2021
project page  /  arXiv  /  code

We introduce a deep multi-view stereo network that jointly models all depth maps in scene space, allowing it to learn geometric priors across entire scenes. It iteratively refines its predictions, leading to highly coherent reconstructions.

VoRTX: Volumetric 3D Reconstruction with Transformers for Voxel-wise View Selection and Fusion
Noah Stier, Alexander Rich, Pradeep Sen, Tobias Höllerer
3DV 2021 (Oral Presentation)
arXiv  /  code

We introduce a 3D reconstruction model with a novel multi-view fusion method based on transformers. It models occlusion by predicting projective occupancy, which reduces noise and leads to more detailed and complete reconstructions.

Using Synthetic Data Generation to Probe Multi-View Stereo Networks
Pranav Acharya, Daniel Lohn, Vivian Ross, Maya Ha, Alexander Rich, Ehsan Sayyad, Tobias Höllerer
ILDAV Workshop, Adjunct Proceedings of ICCV 2021
PDF  /  code

We introduce a synthetic data generation pipeline, which produces image sequences for multi-view stereo methods given 3D scene models as input. We utilize this pipeline to study existing multi-view-stereo depth-prediction methods. Among other results, we find that camera height and vertical camera viewing angle are the parameters that cause the most variation in depth-prediction errors on these image sequences.