Year: 2025
Type: preprint
Abstract: We propose a novel algorithm for accelerating dense long-term 3D point tracking in videos. Through analysis of existing state-of-the-art methods, we identify two major computational bottlenecks. First, transformer-based iterative tracking becomes expensive when handling a large number of trajectories. To address this, we introduce a coarse-to-fine strategy that begins tracking with a small subset of points and progressively expands the set of tracked trajectories. The newly added trajectories are initialized using a learnable interpolation module, which is trained end-to-end alongside the tracking network. Second, we propose an optimization that significantly reduces the cost of correlation feature computation, another key bottleneck in prior methods. Together, these improvements lead to a 5-100x speedup over existing approaches while maintaining state-of-the-art tracking accuracy.
Source: arXiv (Cornell University)
Authors: Ngo, Tuan Duc, Mirzaei, Ashkan, Qian, Guocheng, Liang, Hanwen, Gan, Chuang, Kalogerakis, Evangelos, Wonka, Peter, Wang, Chaoyang
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4417098489
Cites:
Cited by:
Related to: