Research map / Detection and segmentation / Tracking
Structured Siamese Network for Real-Time Visual Tracking
Yunhua Zhang, Lijun Wang, Jinqing Qi, Dong Wang, Mengyang Feng, Huchuan Lu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the integration layer (ablation) | StructSiam · OTB2015 | success AUC: 0.588 -> 0.621 | Improved |
| add the message passing module instead of plain 3x3 conv (ablation) | StructSiam · OTB2015 | success AUC: 0.524 -> 0.621 | Improved |
| add the local pattern detectors (ablation) | StructSiam · OTB2015 | success AUC: 0.531 -> 0.621 | Improved |
| add ALOV300 to ILSVRC VID training data (ablation) | StructSiam · OTB2015 | success AUC: 0.610 -> 0.621 | Improved |
| add the integration layer to SiameseFC (ablation) | SiameseFC · OTB2015 | success AUC: 0.582 -> 0.506 | Hurt |
| use StructSiam instead of SiameseFC | SiameseFC · OTB2015 | success AUC: 0.582 -> 0.621 | Improved |
| use StructSiam instead of SiamFC for tracking speed | SiameseFC · OTB speed | speed (fps): 58 -> 45 | Hurt |
| use StructSiam (CRF message passing) instead of SiamFC under motion blur | SiameseFC · OTB2015 motion blur | success AUC: 0.568 -> 0.642 | Improved |
| use StructSiam instead of SiamFC under occlusion | SiameseFC · OTB2015 occlusion | success AUC: 0.543 -> 0.602 | Improved |
| use ResNet-based SiameseFC instead of StructSiam | StructSiam · VOT2016 | EAO: 0.264 -> 0.277 | Improved |
| use StructSiam instead of SiamAN with same network depth | SiamAN · VOT2016 | EAO: 0.235 -> 0.264 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- Staple — Staple: Complementary Learners for Real-Time Tracking (CVPR 2016) · better on 2 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Learning Dynamic Siamese Network for Visual Object Tracking (ICCV 2017 · similarity 0.89)
- Deeper and Wider Siamese Networks for Real-Time Visual Tracking (CVPR 2019 · similarity 0.89)
- Visual Object Tracking with Discriminative Filters and Siamese Networks: A Survey and Outlook (TPAMI 2022 · similarity 0.89)
- Joint Classification and Regression for Visual Tracking with Fully Convolutional Siamese Networks (IJCV 2022 · similarity 0.89)
- High Performance Visual Tracking With Siamese Region Proposal Network (CVPR 2018 · similarity 0.88)
- A Twofold Siamese Network for Real-Time Object Tracking (CVPR 2018 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
