A Twofold Siamese Network for Real-Time Object Tracking
arXiv:1802.08817
Abstract
Observing that Semantic features learned in an image classification task and Appearance features learned in a similarity matching task complement each other, we build a twofold Siamese network, named SA-Siam, for real-time object tracking. SA-Siam is composed of a semantic branch and an appearance branch. Each branch is a similarity-learning Siamese network. An important design choice in SA-Siam is to separately train the two branches to keep the heterogeneity of the two types of features. In addition, we propose a channel attention mechanism for the semantic branch. Channel-wise weights are computed according to the channel activations around the target position. While the inherited architecture from SiamFC \cite{SiamFC} allows our tracker to operate beyond real-time, the twofold design and the attention mechanism significantly improve the tracking performance. The proposed SA-Siam outperforms all other real-time trackers by a large margin on OTB-2013/50/100 benchmarks.
arXiv commentAccepted by CVPR'18
References in corpus (8)
- TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
- Modeling and Propagating CNNs in a Tree Structure for Visual Tracking
- Siamese Instance Search for Tracking
- Large Margin Object Tracking with Circulant Feature Maps
- Learning Multi-Domain Convolutional Neural Networks for Visual Tracking
- Parallel Tracking and Verifying: A Framework for Real-Time and High Accuracy Visual Tracking
- Learning Policies for Adaptive Tracking with Deep Feature Cascades
- End-to-end Learning of Driving Models from Large-scale Video Datasets
Cited by in corpus (18)
- Fast Online Object Tracking and Segmentation: A Unifying Approach
- Unsupervised Deep Tracking
- Siamese Cascaded Region Proposal Networks for Real-Time Visual Tracking
- Benchmarking Deep Trackers on Aerial Videos
- RGB-T Object Tracking:Benchmark and Baseline
- An In-Depth Analysis of Visual Tracking with Siamese Neural Networks
- Multi-Task Driven Feature Models for Thermal Infrared Tracking
- A Two-Stream Mutual Attention Network for Semi-supervised Biomedical Segmentation with Noisy Labels
- DensSiam: End-to-End Densely-Siamese Network with Self-Attention Model for Object Tracking
- SiamMOT: Siamese Multi-Object Tracking
- High-Performance Long-Term Tracking with Meta-Updater
- Towards a Better Match in Siamese Network Based Visual Object Tracker
- Detect or Track: Towards Cost-Effective Video Object Detection/Tracking
- Model-free Tracking with Deep Appearance and Motion Features Integration
- Protecting Anonymous Speech: A Generative Adversarial Network Methodology for Removing Stylistic Indicators in Text
- 3D Siamese Voxel-to-BEV Tracker for Sparse Point Clouds
- Matching Targets Across Domains with RADON, the Re-Identification Across Domain Network
- Where were my keys? -- Aggregating Spatial-Temporal Instances of Objects for Efficient Retrieval over Long Periods of Time