Fast Tracking via Spatio-Temporal Context Learning
lastmod 2026-11-10created 2017-12-142 edits
Fast Tracking via Spatio-Temporal Context Learning (STC) is a paper proposing a single-object tracking algorithm that models the spatio-temporal context relationship between the tracked target and its surrounding region in a Bayesian framework to estimate the target's location in the next frame. Because computation is done with the FFT, it runs very fast, and thanks to the surrounding context it is relatively robust to occlusion and appearance changes. This page collects the paper and the authors' site.
Local Download
- Fast Tracking via Spatio-Temporal Context Learning
- https://arxiv.org/abs/1311.1939v1
-
1311.1939v1.pdf