ProTracker: Probabilistic Integration for Robust and Accurate Point Tracking
This work addresses the challenge of precise and robust point tracking in videos for computer vision applications, representing an incremental improvement by combining existing paradigms.
The paper tackled the problem of accurate and robust long-term dense point tracking in videos by proposing ProTracker, a probabilistic framework that integrates local optical flow and global heatmaps, achieving state-of-the-art performance among optimization-based approaches and surpassing supervised feed-forward methods on multiple benchmarks.
We propose ProTracker, a novel framework for accurate and robust long-term dense tracking of arbitrary points in videos. Previous methods relying on global cost volumes effectively handle large occlusions and scene changes but lack precision and temporal awareness. In contrast, local iteration-based methods accurately track smoothly transforming scenes but face challenges with occlusions and drift. To address these issues, we propose a probabilistic framework that marries the strengths of both paradigms by leveraging local optical flow for predictions and refined global heatmaps for observations. This design effectively combines global semantic information with temporally aware low-level features, enabling precise and robust long-term tracking of arbitrary points in videos. Extensive experiments demonstrate that ProTracker attains state-of-the-art performance among optimization-based approaches and surpasses supervised feed-forward methods on multiple benchmarks. The code and model will be released after publication.