What Makes Good Synthetic Training Data for Zero-Shot Stereo Matching?
This work addresses the problem of creating effective synthetic training data for stereo matching, which is important for computer vision researchers and practitioners, though it is incremental in nature.
The authors investigated what makes synthetic datasets effective for training stereo matching networks by varying procedural generation parameters, finding that training only on their optimized dataset outperformed training on mixed datasets and was competitive with FoundationStereo while providing open-source code.
Synthetic datasets are a crucial ingredient for training stereo matching networks, but the question of what makes a stereo dataset effective remains underexplored. We investigate the design space of synthetic datasets by varying the parameters of a procedural dataset generator, and report the effects on zero-shot stereo matching performance using standard benchmarks. We validate our findings by collecting the best settings and creating a large-scale dataset. Training only on this dataset achieves better performance than training on a mixture of widely used datasets, and is competitive with training on the FoundationStereo dataset, with the additional benefit of open-source generation code and an accompanying parameter analysis to enable further research. We open-source our system at https://github.com/princeton-vl/InfinigenStereo to enable further research on procedural stereo datasets.