Joint Feature Learning and Relation Modeling for Tracking: A One-Stream Framework

By Botao Ye, Hong Chang, Bingpeng Ma, Shiguang Shan. European Conference on Computer Vision, 2022

973 citations recorded by the source corpus.

The current popular two-stream, two-stage tracking framework extracts the template and the search region features separately and then performs relation modeling, thus the extracted features lack the awareness of the target and have limited target-background discriminability. To tackle the above issue, we propose a novel one-stream tracking (OSTrack) framework that unifies feature learning and relation modeling by bridging the template-search image pairs with bidirectional information flows. In this way, discriminative target-oriented features can be dynamically extracted by mutual guidance. Since no extra heavy relation modeling module is needed and the implementation is highly parallelized, the proposed tracker runs at a fast speed. To further improve the inference efficiency, an in-network candidate early elimination module is proposed based on the strong similarity prior calculated in the one-stream framework. As a unified framework, OSTrack achieves state-of-the-art performance on multiple benchmarks, in particular, it shows impressive results on the one-shot tracking benchmark GOT-10k, i.e., achieving 73.7% AO, improving the existing best result (SwinTrack) by 4.3\%. Besides, our method maintains a good performance-speed trade-off and shows faster convergence. The code and models are available at https://github.com/botaoye/OSTrack.

Subfields: Biomedical Engineering, Computer Vision and Pattern Recognition, Environmental Engineering

View Paper: https://www.semanticscholar.org/paper/c1329f91cfa11011712227c8765fbbe38b9f2b7e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

1. Huchuan Lu
2. Bineng Zhong

In Computer Vision and Pattern Recognition

1. Huchuan Lu
2. Dong Wang
3. Houwen Peng
4. Bineng Zhong
5. Xin Chen
6. Qihua Liang
7. Xianxian Li
8. Jiawen Zhu

Citations in the corpus

• SeqTrack: Sequence to Sequence Learning for Visual Object Tracking (2023, methodology): “2, SeqTrack-B256 performs slightly better than the recent state-of-the-art method OSTrack-256 [55], getting 0.8% AUC score improvement, using the same ViT-B encoder architecture Table 2.”
• Visual Prompt Multi-Modal Tracking (2023, methodology): “Though our ViPT is a short-term algorithm, Table 1 shows ViPT surpasses all previous SOTA trackers and obtains the highest F-score of 59.4%, gaining a significant improvement of 6.5% compared with our foundation model [49].”
• ODTrack: Online Dense Temporal Token Learning for Visual Tracking (2024, methodology): “Video Sequence Sampling Strategy Most existing trackers (Yan et al. 2021a; Cui et al. 2022; Ye et al. 2022) commonly sample image-pairs within a short-term interval, such as 50, 100, or 200 frame intervals.”