mlsanityData to 7 Oct 2026

UHP-SOT: An Unsupervised High-Performance Single Object Tracker

arXiv 2110.01812 ↗First posted 5 Oct 2021cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

An unsupervised online object tracking method that exploits both foreground and background correlations is proposed and named UHP-SOT (Unsupervised High-Performance Single Object Tracker) in this work. UHP-SOT consists of three modules: 1) appearance model update, 2) background motion modeling, and 3) trajectory-based box prediction. A state-of-the-art discriminative correlation filters (DCF) based tracker is adopted by UHP-SOT as the first module. We point out shortcomings of using the first module alone such as failure in recovering from tracking loss and inflexibility in object box adaptation and then propose the second and third modules to overcome them. Both are novel in single object tracking (SOT). We test UHP-SOT on two popular object tracking benchmarks, TB-50 and TB-100, and show that it outperforms all previous unsupervised SOT methods, achieves a performance comparable with the best supervised deep-learning-based SOT methods, and operates at a fast speed (i.e. 22.7-32.0 FPS on a CPU).

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 22 of its 43 references that are in the map. First posted on arXiv 5 Oct 2021.

Builds on

22 of 43 references.

References · 22

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2016Fully-Convolutional Siamese Networks for Object Trackingcited by 566
  3. 2018SiamRPN++: Evolution of Siamese Visual Tracking with Very Deep Networkscited by 445
  4. 2016ECO: Efficient Convolution Operators for Trackingcited by 239
  5. 2015Learning Multi-Domain Convolutional Neural Networks for Visual Trackingcited by 209
  6. 2018Distractor-aware Siamese Networks for Visual Object Trackingcited by 187
  7. 2017Computer Vision for Autonomous Vehicles: Problems, Datasets and State of the Artcited by 182
  8. 2014Return of the Devil in the Details: Delving Deep into Convolutional Netscited by 122
  9. 2017End-to-end representation learning for Correlation Filter based trackingcited by 90
  10. 2016Learning to Track at 100 FPS with Deep Regression Networkscited by 89
  11. 2019Unsupervised Deep Trackingcited by 74
  12. 2016Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Trackingcited by 66
10 more
  1. 2016Learning Spatially Regularized Correlation Filters for Visual Trackingcited by 66
  2. 2018Learning Spatial-Temporal Regularized Correlation Filters for Visual Trackingcited by 63
  3. 2018A Twofold Siamese Network for Real-Time Object Trackingcited by 62
  4. 2015Staple: Complementary Learners for Real-Time Trackingcited by 60
  5. 2016Siamese Instance Search for Trackingcited by 60
  6. 2016Discriminative Scale Space Trackingcited by 46
  7. 2016Discriminative Correlation Filter with Channel and Spatial Reliabilitycited by 45
  8. 2017CREST: Convolutional Residual Learning for Visual Trackingcited by 29
  9. 2018Deep Attentive Tracking via Reciprocative Learningcited by 14
  10. 2016Adaptive Decontamination of the Training Set: A Unified Formulation for Discriminative Visual Trackingcited by 12

Built on by

Cited by year

Citing papers per year: 2021: 1; 2022: 2; 2023: 3.

6 papers.

  1. 2022GUSOT: Green and Unsupervised Single Object Tracking for Long Video Sequencescited by 4
  2. 2021Unsupervised Lightweight Single Object Tracking with UHP-SOT++cited by 3
  3. 2022Acceleration of Subspace Learning Machine via Particle Swarm Optimization and Parallel Processingcited by 2
  4. 2023Green Steganalyzer: A Green Learning Approach to Image Steganalysiscited by 1
  5. 2023A Tiny Machine Learning Model for Point Cloud Object Classificationcited by 0
  6. 2023Unsupervised Green Object Tracker (GOT) without Offline Pre-trainingcited by 0