Staple- Complementary Learners for Real-Time Tracking.pdf

Staple- Complementary Learners for Real-Time Tracking.pdf -...

Info icon This preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Staple: Complementary Learners for Real-Time Tracking Luca Bertinetto Jack Valmadre Stuart Golodetz Ondrej Miksik Philip H.S. Torr University of Oxford { name.surname } @eng.ox.ac.uk ±² />S>S?T >Staple /,?T ±³´ ±µ´ ±°´ ±²´´ ±² colour histogram scores 3HO2G ±²´ ±µ´ Figure 1: Sometimes colour distributions are not enough to discriminate the target from the background. Conversely, template models (like HOG) depend on the spatial configuration of the object and perform poorly when this changes rapidly. Our tracker Staple can rely on the strengths of both template and colour-based models. Like DSST [ 10 ], its performance is not affected by non-distinctive colours (top). Like DAT [ 33 ], it is robust to fast deformations (bottom). Abstract Correlation Filter-based trackers have recently achieved excellent performance, showing great robustness to chal- lenging situations exhibiting motion blur and illumination changes. However, since the model that they learn depends strongly on the spatial layout of the tracked object, they are notoriously sensitive to deformation. Models based on colour statistics have complementary traits: they cope well with variation in shape, but suffer when illumination is not consistent throughout a sequence. Moreover, colour distri- butions alone can be insufficiently discriminative. In this paper, we show that a simple tracker combining comple- mentary cues in a ridge regression framework can operate faster than 80 FPS and outperform not only all entries in the popular VOT14 competition, but also recent and far more sophisticated trackers according to multiple benchmarks. 1. Introduction We consider the widely-adopted scenario of short-term, single-object tracking, in which the target is only specified in the first frame (using a rectangle). Short-term implies that re-detection should not be necessary. The key challenge of tracking an unfamiliar object in video is to be robust to changes in its appearance. The task of tracking unfamil- iar objects, for which training examples are not available in advance, is interesting because in many situations it is not feasible to obtain such a dataset. It is advantageous for the algorithm to perform above real-time for computationally intensive applications such as robotics, surveillance, video processing and augmented reality. Since an object’s appearance can vary significantly dur- ing a video, it is not generally effective to estimate its model from the first frame alone and use this single, fixed model to locate the object in all other frames. Most state-of-the- art algorithms therefore employ model adaptation to take advantage of information present in later frames. The sim- plest, most widespread approach is to treat the tracker’s pre- dictions in new frames as training data with which to update the model. The danger of learning from predictions is that small errors can accumulate and cause model drift. This is particularly likely to happen when the appearance of the object changes.
Image of page 1

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern