mlsanityData to 7 Oct 2026

CHOSEN: Contrastive Hypothesis Selection for Multi-View Depth Refinement

arXiv 2404.02225 ↗First posted 2 Apr 2024cs.CV cs.AIcited by 4

Not on the map. Why?

Abstract (arXiv)

We propose CHOSEN, a simple yet flexible, robust and effective multi-view depth refinement framework. It can be employed in any existing multi-view stereo pipeline, with straightforward generalization capability for different multi-view capture systems such as camera relative positioning and lenses. Given an initial depth estimation, CHOSEN iteratively re-samples and selects the best hypotheses, and automatically adapts to different metric or intrinsic scales determined by the capture system. The key to our approach is the application of contrastive learning in an appropriate solution space and a carefully designed hypothesis feature, based on which positive and negative hypotheses can be effectively distinguished. Integrated in a simple baseline multi-view stereo pipeline, CHOSEN delivers impressive quality in terms of depth and normal accuracy compared to many current deep learning based multi-view stereo pipelines.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 29 of its 53 references that are in the map. First posted on arXiv 2 Apr 2024.

Builds on

29 of 53 references.

References · 29

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  2. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,764
  3. 2020RAFT: Recurrent All-Pairs Field Transforms for Optical Flowcited by 2,385
  4. 2018MVSNet: Depth Inference for Unstructured Multi-view Stereocited by 780
  5. 2019BlendedMVS: A Large-scale Dataset for Generalized Multi-view Stereo Networkscited by 487
  6. 2018Pyramid Stereo Matching Networkcited by 486
  7. 2019Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matchingcited by 414
  8. 2017End-to-End Learning of Geometry and Context for Deep Stereo Regressioncited by 359
  9. 2019Recurrent MVSNet for High-resolution Multi-view Stereo Depth Inferencecited by 252
  10. 2020PatchmatchNet: Learned Multi-View Patchmatch Stereocited by 204
  11. 2017Learning a Multi-View Stereo Machinecited by 194
  12. 2019Deep Stereo using Adaptive Thin Volume Representation with Uncertainty Awarenesscited by 151
17 more
  1. 2021NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in the Wildcited by 136
  2. 2017Learning Efficient Point Cloud Generation for Dense 3D Object Reconstructioncited by 133
  3. 2017SurfaceNet: An End-to-end 3D Neural Network for Multiview Stereopsiscited by 128
  4. 2020HITNet: Hierarchical Iterative Tile Refinement Network for Real-time Stereo Matchingcited by 120
  5. 2021TransMVSNet: Global Context-aware Multi-view Stereo Network with Transformerscited by 117
  6. 2020Fast-MVSNet: Sparse-to-Dense Multi-View Stereo With Learned Propagation and Gauss-Newton Refinementcited by 105
  7. 2018Unsupervised Learning of Shape and Pose with Differentiable Point Cloudscited by 89
  8. 2022Multiface: A Dataset for Neural Face Renderingcited by 87
  9. 2015The Fast Bilateral Solvercited by 86
  10. 2022Rethinking Depth Estimation for Multi-View Stereo: A Unified Representationcited by 82
  11. 2017Cascade Residual Learning: A Two-stage Convolutional Neural Network for Stereo Matchingcited by 63
  12. 2022MVSFormer: Multi-View Stereo by Learning Robust Image Features and Temperature-based Depthcited by 50
  13. 2024MVSFormer++: Revealing the Devil in Transformer's Details for Multi-View Stereocited by 49
  14. 2018ActiveStereoNet: End-to-End Self-Supervised Learning for Active Stereo Systemscited by 35
  15. 2020PVSNet: Pixelwise Visibility-Aware Multi-View Stereo Networkcited by 35
  16. 2021Generalized Binary Search Network for Highly-Efficient Multi-View Stereocited by 32
  17. 2019DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstructioncited by 23

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 3.

4 papers.

  1. 2025Pixels2Points: Fusing 2D and 3D Features for Facial Skin Segmentationcited by 2
  2. 2026Representing 3D Faces with Learnable B-Spline Volumescited by 0
  3. 2026VEPHand: View-Efficient Photometric Hand Performance Capture at Scalecited by 0
  4. 2026GNM Head: A Generative aNthropometric Model of the human headcited by 0