Flow-Anything: Learning Real-World Optical Flow Estimation from Large-Scale Single-view Images
This addresses the domain gap issue for computer vision researchers and practitioners by enabling more robust optical flow estimation from real-world images, though it is incremental as it builds on existing monocular depth estimation and rendering techniques.
The paper tackled the problem of limited real-world robustness in optical flow estimation due to reliance on synthetic datasets by proposing Flow-Anything, a framework that generates optical flow training data from large-scale single-view images, resulting in models that outperform advanced unsupervised and supervised methods on synthetic datasets.
Optical flow estimation is a crucial subfield of computer vision, serving as a foundation for video tasks. However, the real-world robustness is limited by animated synthetic datasets for training. This introduces domain gaps when applied to real-world applications and limits the benefits of scaling up datasets. To address these challenges, we propose \textbf{Flow-Anything}, a large-scale data generation framework designed to learn optical flow estimation from any single-view images in the real world. We employ two effective steps to make data scaling-up promising. First, we convert a single-view image into a 3D representation using advanced monocular depth estimation networks. This allows us to render optical flow and novel view images under a virtual camera. Second, we develop an Object-Independent Volume Rendering module and a Depth-Aware Inpainting module to model the dynamic objects in the 3D representation. These two steps allow us to generate realistic datasets for training from large-scale single-view images, namely \textbf{FA-Flow Dataset}. For the first time, we demonstrate the benefits of generating optical flow training data from large-scale real-world images, outperforming the most advanced unsupervised methods and supervised methods on synthetic datasets. Moreover, our models serve as a foundation model and enhance the performance of various downstream video tasks.