Research map / Detection and segmentation / Tracking
Visual Tracking via Dynamic Memory Networks
Tianyu Yang, Antoni B. Chan
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 distractor template canceling and auxiliary classification loss (MemDTC*) to MemTrack | MemTrack · OTB-2015 | DP@20 (%): 82.0 -> 84.8 | Improved |
| add distractor template canceling and auxiliary loss (MemDTC*) to MemTrack | MemTrack · OTB-2013 | CLE (pixel): 27.6 -> 16.9 | Improved |
| add distractor template canceling (MemDTC) to MemTrack (ablation) | MemTrack · OTB-2013/2015 | precision and success plots | Improved |
| add auxiliary classification loss to MemDTC (MemDTC* vs MemDTC) (ablation) | MemDTC · VOT-2015 | EAO: 0.2948 -> 0.3005 | Improved |
| add auxiliary classification loss to MemTrack (MemTrack*) (ablation) | MemTrack · VOT-2016 | EAO: 0.2723 -> 0.2713 | Hurt |
| add auxiliary classification loss to MemTrack (MemTrack*) (ablation) | MemTrack · VOT-2017 | EAO: 0.2427 -> 0.2416 | Hurt |
| add auxiliary loss to MemDTC (MemDTC* vs MemDTC) on OTB-2015 vs OTB-2013 (ablation) | MemTrack · OTB-2013 and OTB-2015 | precision/success | Mixed |
| add attention mechanism (vs averaging feature vectors) (ablation) | MemTrack · OTB-2015 | success/precision | Improved |
| add dynamic memory network (vs queue memory with averaged templates) (ablation) | MemTrack · OTB-2015 | success/precision | Improved |
| use soft weighted reading instead of hard reading (ablation) | MemTrack · OTB-2015 | success/precision | Improved |
| add channel-wise gated residual template learning (ablation) | MemTrack · OTB-2015 | success/precision | Improved |
| increase positive memory slots (ablation) | MemTrack · OTB-2015 | success plot: saturates at 8 memory slots | Improved |
Closest papers by meaning
- Learning Dynamic Memory Networks for Object Tracking (ECCV 2018 · similarity 0.98)
- STMTrack: Template-Free Visual Tracking With Space-Time Memory Networks (CVPR 2021 · similarity 0.89)
- Reading Relevant Feature from Global Representation Memory for Visual Object Tracking (NeurIPS 2023 · similarity 0.89)
- MUTrack: A Memory-Aware Unified Representation Framework for Visual Tracking (AAAI 2026 · similarity 0.88)
- A Distractor-Aware Memory for Visual Object Tracking with SAM2 (CVPR 2025 · similarity 0.86)
- MIMTrack: In-Context Tracking via Masked Image Modeling (AAAI 2025 · similarity 0.86)
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.
