Zoom Link: https://cmu.zoom.us/j/5523238059

Title: Robust Instance Tracking via Uncertainty Flow

Abstract:

Current state-of-the-art trackers often fail due to distractors and large object appearance changes. In this work, we explore the use of dense optical flow to improve tracking robustness. Our main insight is that, because flow estimation can also have errors, we need to incorporate an estimate of flow uncertainty for robust tracking. We present a novel tracking framework which combines appearance and flow uncertainty information to track objects in challenging scenarios. We experimentally verify that our framework improves tracking robustness, leading to new state-of-the-art results. Further, our  experimental ablations show the importance of flow uncertainty for robust tracking.

Committee:

David Held (advisor)

Deva Ramanan
Leo Keselman