Abstract:Visual tracking addresses the problem of identifying and localizing an unknown target in a video given the target specified by a bounding box in the first frame. In this paper, we propose a dual network to better utilize features among layers for visual tracking. It is observed that features in higher layers encode semantic context while its counterparts in lower layers are sensitive to discriminative appearance. Thus we exploit the hierarchical features in different layers of a deep model and design a dual structure to obtain better feature representation from various streams, which is rarely investigated in previous work. To highlight geometric contours of the target, we integrate the hierarchical feature maps with an edge detector as the coarse prior maps to further embed local details around the target. To leverage the robustness of our dual network, we train it with random patches measuring the similarities between the network activation and target appearance, which serves as a regularization to enforce the dual network to focus on target object. The proposed dual network is updated online in a unique manner based on the observation that the target being tracked in consecutive frames should share more similar feature representations than those in the surrounding background. It is also found that for a target object, the prior maps can help further enhance performance by passing message into the output maps of the dual network. Therefore, an independent component analysis with reference algorithm (ICA-R) is employed to extract target context using prior maps as guidance. Online tracking is conducted by maximizing the posterior estimate on the final maps with stochastic and periodic update. Quantitative and qualitative evaluations on two large-scale benchmark data sets show that the proposed algorithm performs favourably against the state-of-the-arts.

Single Online Visual Object Tracking with Enhanced Tracking and Detection Learning

Visual Tracking Jointly with Online and Offline Learning

Occlusion-Aware Real-Time Object Tracking

Visual Tracking with Long-Short Term Based Correlation Filter

Survey of Single-Target Visual Tracking Methods Based on Online Learning.

Single Object Tracking Research: A Survey

Recent Advances of Single-Object Tracking Methods: A Brief Survey.

Real-Time Online Multi-Object Tracking

Discriminative and Robust Online Learning for Siamese Visual Tracking

Online Unsupervised Feature Learning for Visual Tracking

Robust and Real-Time Visual Tracking Based on Complementary Learners.

Dual Deep Network for Visual Tracking

Online dual dictionary learning for visual object tracking

Describe and Attend to Track: Learning Natural Language guided Structural Representation and Visual Attention for Object Tracking

Online Visual Multi-Object Tracking via Labeled Random Finite Set Filtering

Distractor-aware discrimination learning for online multiple object tracking

Cascaded Iterative Training Model and Parallel Multi-Classifiers for Visual Object Tracking

Online multi-object tracking with pedestrian re-identification and occlusion processing

Visual Tracking via Sparse Representation and Online Dictionary Learning.

Learning a Deep Compact Image Representation for Visual Tracking

DeepTrack: Learning Discriminative Feature Representations Online for Robust Visual Tracking