High Performance Visual Tracking with Siamese Region Proposal Network
By Bo Li, Junjie Yan, Wei Wu, Zheng Zhu, Xiaolin Hu. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018
2,552 citations recorded by the source corpus.
Visual object tracking has been a fundamental topic in recent years and many deep learning based trackers have achieved state-of-the-art performance on multiple benchmarks. However, most of these trackers can hardly get top performance with real-time speed. In this paper, we propose the Siamese region proposal network (Siamese-RPN) which is end-to-end trained off-line with large-scale image pairs. Specifically, it consists of Siamese subnetwork for feature extraction and region proposal subnetwork including the classification branch and regression branch. In the inference phase, the proposed framework is formulated as a local one-shot detection task. We can pre-compute the template branch of the Siamese subnetwork and formulate the correlation layers as trivial convolution layers to perform online tracking. Benefit from the proposal refinement, traditional multi-scale test and online fine-tuning can be discarded. The Siamese-RPN runs at 160 FPS while achieving leading performance in VOT2015, VOT2016 and VOT2017 real-time challenges.
Subfields: Computer Vision and Pattern Recognition, Safety, Risk, Reliability and Quality
View Paper: https://www.semanticscholar.org/paper/320d05db95ab42ade69294abe46cd1aca6aca602
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Ming-Hsuan Yang
- Dacheng Tao
- Xiaogang Wang
- Trevor Darrell
- Yi Yang
- Chunhua Shen
- Andrea Vedaldi
In Safety, Risk, Reliability and Quality
Citations in the corpus
- MOTChallenge: A Benchmark for Single-Camera Multiple Target Tracking (2020, methodology): “formance of submitted models over time and by type. The main common components of top-performing methods are: (i) learned single-target regressors (singleobject trackers), such as [Held et al., 2016, Li et al., 2018], and (ii) re-identication modules [Bergmann et al., 2019]. These methods fall short”
- Backbone is All Your Need: A Simplified Architecture for Visual Object Tracking (2022, methodology): “In stead of directly getting the target position through the response map, SiamRPN [27] and the follow-ups [59,18,11,52] send the response map to Region Proposal Network (RPN) [38] to get a more accurate localization and scale estimation.”
- Siamese Box Adaptive Network for Visual Tracking (2020, methodology): “Although anchor-based trackers [21, 39, 20] can handle changes in scale and aspect ratio, it is necessary to carefully design and fix the parameters of the anchor boxes.”
- AiATrack: Attention in Attention for Transformer Visual Tracking (2022, background): “Over the past few years, Siamese trackers [2,35,34,67], which regards the visual tracking task as a one-shot matching problem, have gained enormous popularity.”
- Deeper and Wider Siamese Networks for Real-Time Visual Tracking (2019, methodology): “0%(EAO) relative improvements over the original versions [2, 20] on the OTB-15, VOT-16 and VOT-17 datasets, respectively.”
- Graph Attention Tracking (2020, positive): “Existing correlation based information embedding meth-ods [2, 16, 15] take the whole target feature as a unity to match with the search features.”
- SeqTrack: Sequence to Sequence Learning for Visual Object Tracking (2023, methodology): “Prior knowledge, such as window penalty [2,27] and online update [12, 28, 53], has been widely incorporated into existing tracking models and proved effective [7,10,43,55].”
- Distractor-aware Siamese Networks for Visual Object Tracking (2018, methodology): “The modified AlexNet [15] pretrained using ImageNet [28] is used as described in SiamRPN [16].”
- Towards More Flexible and Accurate Object Tracking with Natural Language: Algorithms and Benchmark (2021, positive): “When we take the result of visual grounding in the first frame as the initialized bbox of visual tracker SiamRPN++, we achieve 0.24|0.19 on the OTB-Lang dataset.”
- Learning Correspondence From the Cycle-Consistency of Time (2019, background): “[49], researchers largely turned to “tracking as repeated recognition”, where trained object detectors are applied to each frame independently [3, 28, 80, 71, 19, 35, 65].”
- The Seventh Visual Object Tracking VOT2019 Challenge Results (2019, background): “In a second step, SiamRPN++ predicts the accurate position and the size of the target.”
- SwinTrack: A Simple and Strong Baseline for Transformer Tracking (2021, methodology): “In order to improve [1] in dealing with scale variation, the method of [26] incorporates the region proposal network into Siamese network and proposes the anchor-based tracker, achieving higher accuracy with faster speed.”
- Learning Target Candidate Association to Keep Track of What Not to Track (2021, background): “The two currently dominating paradigms are Siamese networks [2, 35, 34] and discriminative appearance modules [3, 13].”
- Learning Discriminative Model Prediction for Tracking (2019, methodology): “Recently, methods based on Siamese networks [2, 22, 38] have received much attention due to their end-to-end training capabilities and high efficiency.”
- Unsupervised Deep Tracking (2019, methodology): “The SiamRPN [27] additionally uses more labeled training videos from Youtube-BB dataset [41].”
- Fast Online Object Tracking and Segmentation: A Unifying Approach (2018, methodology): “[31] considerably improve the performance of SiamFC by relying on a region proposal network (RPN) [53, 16], which allows to estimate the target location with a bounding box of variable aspect ratio.”
- ODTrack: Online Dense Temporal Token Learning for Visual Tracking (2024, methodology): “To improve the accuracy and robustness of trackers, several different approaches are proposed, such as prediction head networks (Li et al. 2018; Chen et al. 2020; Zhang et al. 2020), cross-correlation modules (Han et al. 2021; Liao et al. 2020; Chen et al. 2021), powerful backbone (Chen et al.…”