Research map / Detection and segmentation / CNN features
Bridging the Gap Between Detection and Tracking: A Unified Approach
Lianghua Huang, Xin Zhao, Kaiqi Huang
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 |
|---|---|---|---|
| use learned few-shot (MAML) finetuning instead of brute-force gradient descent (ablation) | SSD tracker · OTB-2013 | AUC: 0.551 -> 0.637 | Improved |
| use learned few-shot (MAML) finetuning instead of brute-force gradient descent (ablation) | SSD tracker · OTB-2013 | OP: 0.698 -> 0.813 | Improved |
| add anchored updating (ablation) | SSD tracker · OTB-2013 | AUC: 0.612 -> 0.637 | Improved |
| add anchored updating (ablation) | SSD tracker · OTB-2013 | OP: 0.780 -> 0.813 | Improved |
| finetune bounding box regression head as well as classification head (ablation) | SSD tracker · OTB-2013 | AUC: 0.609 -> 0.637 | Improved |
| finetune bounding box regression head as well as classification head (ablation) | SSD tracker · OTB-2013 | OP: 0.773 -> 0.813 | Improved |
| use multi-resolution guidance images matched to layer receptive fields (ablation) | SSD tracker · OTB-2013 | AUC: 0.629 -> 0.637 | Improved |
| use multi-resolution guidance images matched to layer receptive fields (ablation) | SSD tracker · OTB-2013 | OP: 0.807 -> 0.813 | Improved |
| deepen backbone from VGG-16 to ResNet-50 (ablation) | FasterRCNN tracker · OTB-2013 | AUC: 0.639 -> 0.656 | Improved |
| deepen backbone from ResNet-50 to ResNet-101 (ablation) | FasterRCNN tracker · OTB-2013 | AUC: 0.656 -> 0.642 | No clear effect |
| deepen backbone from ResNet-50 to ResNet-101 (ablation) | FasterRCNN tracker · OTB-2013 | OP: 0.829 -> 0.825 | No clear effect |
| replace SSD detector with FasterRCNN (ResNet-50) as base | Detector-based tracker · UAV123 | AUC: 0.531 -> 0.586 | Improved |
Closest papers by meaning
- Tracking Without Bells and Whistles (ICCV 2019 · similarity 0.90)
- TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild (ECCV 2018 · similarity 0.90)
- Beyond Local Search: Tracking Objects Everywhere With Instance-Specific Proposals (CVPR 2016 · similarity 0.89)
- Detect to Track and Track to Detect (ICCV 2017 · similarity 0.88)
- Tracking by Instance Detection: A Meta-Learning Approach (CVPR 2020 · similarity 0.88)
- DiffusionTrack: Diffusion Model for Multi-Object Tracking (AAAI 2024 · similarity 0.87)
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.
