Back to Search
Start Over
Model Decay in Long-Term Tracking
- Source :
- ICPR, Proceedings of ICPR 2020: 25th International Conference on Pattern Recognition : Milan, 10-15 January 2021, 2685-2692, STARTPAGE=2685;ENDPAGE=2692;TITLE=Proceedings of ICPR 2020
- Publication Year :
- 2021
- Publisher :
- IEEE, 2021.
-
Abstract
- To account for appearance variations, tracking models need to be updated during the course of inference. However, updating the tracker model with adverse bounding box predictions adds an unavoidable bias term to the learning. This bias term, which we refer to as model decay, offsets the learning and causes tracking drift. While its adverse affect might not be visible in short-term tracking, accumulation of this bias over a long-term can eventually lead to a permanent loss of the target. In this paper, we look at the problem of model bias from a mathematical perspective. Further, we briefly examine the effect of various sources of tracking error on model decay, using a correlation filter (ECO) and a Siamese (SINT) tracker. Based on observations and insights, we propose simple additions that help to reduce model decay in long-term tracking. The proposed tracker is evaluated on four long-term and one short-term tracking benchmarks, demonstrating superior accuracy and robustness, even on 30 minute long videos.
- Subjects :
- FOS: Computer and information sciences
business.industry
Computer science
Computer Vision and Pattern Recognition (cs.CV)
Perspective (graphical)
Computer Science - Computer Vision and Pattern Recognition
Inference
02 engineering and technology
Tracking (particle physics)
Term (time)
Tracking error
Robustness (computer science)
Minimum bounding box
Pattern recognition (psychology)
0202 electrical engineering, electronic engineering, information engineering
020201 artificial intelligence & image processing
Artificial intelligence
business
Algorithm
Subjects
Details
- Database :
- OpenAIRE
- Journal :
- 2020 25th International Conference on Pattern Recognition (ICPR)
- Accession number :
- edsair.doi.dedup.....9dde9346c696f12b918daf48ff869d00