MegaFlow: Zero-Shot Large Displacement Optical Flow

2026-03-26

AI Summary

This work addresses the challenges of accuracy and generalization in zero-shot optical flow estimation under large displacements by introducing a novel paradigm based on global matching. Departing from task-specific architectures, the method directly leverages globally-aware features extracted from a pretrained Vision Transformer to model large-motion correspondence, followed by a lightweight iterative refinement module to achieve sub-pixel accuracy. This framework represents the first successful transfer of pretrained visual priors to zero-shot optical flow estimation, establishing a unified and generalizable motion modeling pipeline. It achieves state-of-the-art zero-shot performance across multiple optical flow benchmarks and demonstrates exceptional cross-domain generalization capabilities in long-range point tracking tasks.

Technology Category

• Computer Vision: Large Vision Models
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Accurate estimation of large displacement optical flow remains a critical challenge. Existing methods typically rely on iterative local search or/and domain-specific fine-tuning, which severely limits their performance in large displacement and zero-shot generalization scenarios. To overcome this, we introduce MegaFlow, a simple yet powerful model for zero-shot large displacement optical flow. Rather than relying on highly complex, task-specific architectural designs, MegaFlow adapts powerful pre-trained vision priors to produce temporally consistent motion fields. In particular, we formulate flow estimation as a global matching problem by leveraging pre-trained global Vision Transformer features, which naturally capture large displacements. This is followed by a few lightweight iterative refinements to further improve the sub-pixel accuracy. Extensive experiments demonstrate that MegaFlow achieves state-of-the-art zero-shot performance across multiple optical flow benchmarks. Moreover, our model also delivers highly competitive zero-shot performance on long-range point tracking benchmarks, demonstrating its robust transferability and suggesting a unified paradigm for generalizable motion estimation. Our project page is at: https://kristen-z.github.io/projects/megaflow.

Problem

• optical flow
• large displacement
• zero-shot
• motion estimation
• generalization

Innovation

• zero-shot optical flow
• large displacement
• Vision Transformer
• global matching
• motion estimation