# MAGIC: Learning from Visibility Asymmetry for Unsupervised Stereo Matching

arXiv: https://arxiv.org/abs/2508.10838 · HTML: https://mlsanity.com/paper/2508.10838/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 33 of its 51 references that are in the map.
- First posted on arXiv 14 Aug 2025.

## Builds on (most cited first, 30 of 33)

- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Bootstrap your own latent: A new approach to self-supervised Learning](https://mlsanity.com/paper/2006.07733/) (2020)
- [CARLA: An Open Urban Driving Simulator](https://mlsanity.com/paper/1711.03938/) (2017)
- [Depth Anything V2](https://mlsanity.com/paper/2406.09414/) (2024)
- [Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data](https://mlsanity.com/paper/2401.10891/) (2024)
- [A Closer Look at Memorization in Deep Networks](https://mlsanity.com/paper/1706.05394/) (2017)
- [A Large Dataset to Train Convolutional Networks for Disparity, Optical Flow, and Scene Flow Estimation](https://mlsanity.com/paper/1512.02134/) (2015)
- [Pyramid Stereo Matching Network](https://mlsanity.com/paper/1803.08669/) (2018)
- [End-to-End Learning of Geometry and Context for Deep Stereo Regression](https://mlsanity.com/paper/1703.04309/) (2017)
- [RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching](https://mlsanity.com/paper/2109.07547/) (2021)
- [Group-wise Correlation Stereo Network](https://mlsanity.com/paper/1903.04025/) (2019)
- [A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay](https://mlsanity.com/paper/1803.09820/) (2018)
- [Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlation](https://mlsanity.com/paper/2203.11483/) (2022)
- [Iterative Geometry Encoding Volume for Stereo Matching](https://mlsanity.com/paper/2303.06615/) (2023)
- [FoundationStereo: Zero-Shot Stereo Matching](https://mlsanity.com/paper/2501.09898/) (2025)
- [CFNet: Cascade and Fused Cost Volume for Robust Stereo Matching](https://mlsanity.com/paper/2104.04314/) (2021)
- [Selective-Stereo: Adaptive Frequency Information Selection for Stereo Matching](https://mlsanity.com/paper/2403.00486/) (2024)
- [Real-time self-adaptive deep stereo](https://mlsanity.com/paper/1810.05424/) (2018)
- [DEFOM-Stereo: Depth Foundation Model Based Stereo Matching](https://mlsanity.com/paper/2501.09466/) (2025)
- [NeRF-Supervised Deep Stereo](https://mlsanity.com/paper/2303.17603/) (2023)
- [Parallax Attention for Unsupervised Stereo Correspondence Learning](https://mlsanity.com/paper/2009.08250/) (2020)
- [Flow2Stereo: Effective Self-Supervised Learning of Optical Flow and Stereo Matching](https://mlsanity.com/paper/2004.02138/) (2020)
- [Reversing the cycle: self-supervised deep stereo through enhanced monocular distillation](https://mlsanity.com/paper/2008.07130/) (2020)
- [Self-Supervised Learning for Stereo Matching with Self-Improving Ability](https://mlsanity.com/paper/1709.00930/) (2017)
- [AdaStereo: A Simple and Efficient Approach for Adaptive Stereo Matching](https://mlsanity.com/paper/2004.04627/) (2020)
- [Zoom and Learn: Generalizing Deep Stereo Matching to Novel Domains](https://mlsanity.com/paper/1803.06641/) (2018)
- [Adaptive Multi-Modal Cross-Entropy Loss for Stereo Matching](https://mlsanity.com/paper/2306.15612/) (2023)
- [KD-MVS: Knowledge Distillation Based Self-supervised Learning for Multi-view Stereo](https://mlsanity.com/paper/2207.10425/) (2022)
- [Robust Synthetic-to-Real Transfer for Stereo Matching](https://mlsanity.com/paper/2403.07705/) (2024)
- [Stereo Matching by Self-supervision of Multiscopic Vision](https://mlsanity.com/paper/2104.04170/) (2021)

## Abstract (arXiv)

Learning disparity in occluded regions remains difficult for unsupervised stereo matching. Photometric supervision lacks valid target-view correspondences in these regions, while the teacher and student in conventional binocular self-training share the same target view and therefore the same occlusions. Even when supervision is available, the small proportion of occluded pixels limits their contribution to training. We propose MAGIC, a multi-baseline geometric consistency framework for reliable and effective occlusion supervision. The teacher and student share a reference image but use different target views, allowing the teacher to observe correspondences that are occluded from the student. After aligning disparities across baselines, MAGIC uses predictions from teacher-visible regions to supervise student-occluded regions. An occlusion-aware weighting strategy strengthens supervision on teacher-visible but student-occluded pixels, preventing their training signal from being overwhelmed by non-occluded regions. We also introduce MBS20K, a synthetic multi-baseline stereo dataset spanning diverse scenes, weather, and lighting. Pre-trained on MBS20K, MAGIC generalizes to real-world datasets with consistently fewer occluded-region outliers. On KITTI, the pre-trained model already outperforms several fine-tuned unsupervised methods. Fine-tuning this model on standard binocular pairs achieves state-of-the-art unsupervised performance on KITTI 2015 and 2012. Incorporating synthesized multi-baseline views during fine-tuning further improves performance. Our code and dataset will be released upon acceptance.