Fast Online Object Tracking and Segmentation: A Unifying Approach

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 the same time demonstrating competitive performance and the best speed for the semi-supervised video object segmentation task on DAVIS-2016 and DAVIS-2017. The project website is http://www.robots.ox.ac.uk/~qwang/SiamMask.

arXiv commentCVPR 2019 camera ready. Code available at https://github.com/foolwood/SiamMask

References in corpus (15)

• Deep Residual Learning for Image Recognition
• An Overview of Multi-Task Learning in Deep Neural Networks
• Fully Convolutional Networks for Semantic Segmentation
• A Novel Performance Evaluation Methodology for Single-Target Trackers
• Learning feed-forward one-shot learners
• Siamese Instance Search for Tracking
• Distractor-aware Siamese Networks for Visual Object Tracking
• Learning Video Object Segmentation from Static Images
• A Twofold Siamese Network for Real-Time Object Tracking
• TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild
• Learning Dynamic Memory Networks for Object Tracking
• Video Propagation Networks
• Long-term Tracking in the Wild: A Benchmark
• Learning Spatial-Aware Regressions for Visual Tracking
• Towards a Better Match in Siamese Network Based Visual Object Tracker

Cited by in corpus (8)

• Deep Learning for Visual Tracking: A Comprehensive Survey
• Video Object Segmentation and Tracking: A Survey
• Tracking Holistic Object Representations
• Prediction-Tracking-Segmentation
• Self-Supervised Object-in-Gripper Segmentation from Robotic Motions
• Towards Early Prediction of Human iPSC Reprogramming Success
• Occlusion Aware Kernel Correlation Filter Tracker using RGB-D
• Motion Prediction in Visual Object Tracking