Fast Online Object Tracking and Segmentation: A Unifying Approach
Science Oxford · Institute of Automation · +2 more institutions
Abstract
In this paper we illustrate how to perform both visual object tracking and semi-supervised video object segmentation, in real-time, with a single simple approach. Our method, dubbed SiamMask, improves the offline training procedure of popular fully-convolutional Siamese approaches for object tracking by augmenting their loss with a binary segmentation task. Once trained, SiamMask solely relies on a single bounding box initialisation and operates online, producing class-agnostic object segmentation masks and rotated bounding boxes at 55 frames per second. Despite its simplicity, versatility and fast speed, our strategy allows us to establish a new state-of-the-art among real-time trackers on VOT-2018, while at…
Citation impact
- FWCI
- 95.02
- Percentile
- 100%
- References
- 102
Authors
5Topics & keywords
- Computer science
- Artificial intelligence
- Segmentation
- Minimum bounding box
- Computer vision
- Video tracking
- Object (grammar)
- Bounding overwatch