Tracking Everything Everywhere All at Once

Autor:	Wang, Qianqian, Chang, Yen-Yu, Cai, Ruojin, Li, Zhengqi, Hariharan, Bharath, Holynski, Aleksander, Snavely, Noah
Rok vydání:	2023
Předmět:	Computer Science - Computer Vision and Pattern Recognition
Druh dokumentu:	Working Paper
Popis:	We present a new test-time optimization method for estimating dense and long-range motion from a video sequence. Prior optical flow or particle video tracking algorithms typically operate within limited temporal windows, struggling to track through occlusions and maintain global consistency of estimated motion trajectories. We propose a complete and globally consistent motion representation, dubbed OmniMotion, that allows for accurate, full-length motion estimation of every pixel in a video. OmniMotion represents a video using a quasi-3D canonical volume and performs pixel-wise tracking via bijections between local and canonical space. This representation allows us to ensure global consistency, track through occlusions, and model any combination of camera and object motion. Extensive evaluations on the TAP-Vid benchmark and real-world footage show that our approach outperforms prior state-of-the-art methods by a large margin both quantitatively and qualitatively. See our project page for more results: http://omnimotion.github.io/ Comment: ICCV 2023
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2306.05422 Zobrazit plný text záznamu View this record from Arxiv