ZeroStereo: Zero-shot Stereo Matching from Single Images
This addresses the problem of limited annotated real-world stereo data for researchers and practitioners in computer vision, offering an incremental improvement through novel synthesis techniques.
The paper tackles the challenge of generalizing stereo matching to real-world scenarios by proposing ZeroStereo, a pipeline that synthesizes stereo images from single images using pseudo disparities and diffusion inpainting, achieving state-of-the-art zero-shot generalization across multiple datasets with a dataset volume comparable to Scene Flow.
State-of-the-art supervised stereo matching methods have achieved remarkable performance on various benchmarks. However, their generalization to real-world scenarios remains challenging due to the scarcity of annotated real-world stereo data. In this paper, we propose ZeroStereo, a novel stereo image generation pipeline for zero-shot stereo matching. Our approach synthesizes high-quality right images from arbitrary single images by leveraging pseudo disparities generated by a monocular depth estimation model. Unlike previous methods that address occluded regions by filling missing areas with neighboring pixels or random backgrounds, we fine-tune a diffusion inpainting model to recover missing details while preserving semantic structure. Additionally, we propose Training-Free Confidence Generation, which mitigates the impact of unreliable pseudo labels without additional training, and Adaptive Disparity Selection, which ensures a diverse and realistic disparity distribution while preventing excessive occlusion and foreground distortion. Experiments demonstrate that models trained with our pipeline achieve state-of-the-art zero-shot generalization across multiple datasets with only a dataset volume comparable to Scene Flow. Code: https://github.com/Windsrain/ZeroStereo.