Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Tracking
Preprint
Abstract
Discriminative Correlation Filters (DCF) have demonstrated excellent performance for visual object tracking. The key to their success is the ability to efficiently exploit available negative data by including all shifted versions of a training sample. However, the underlying DCF formulation is restricted to single-resolution feature maps, significantly limiting its potential. In this paper, we go beyond the conventional DCF framework and introduce a novel formulation for training continuous convolution filters. We employ an implicit interpolation model to pose the learning problem in the continuous spatial domain. Our proposed formulation enables efficient integration of multi-resolution deep feature maps, leading to superior results on three object tracking benchmarks: OTB-2015 (+5.1% in mean OP), Temple-Color (+4.6% in mean OP), and VOT2015 (20% relative reduction in failure rate). Additionally, our approach is capable of sub-pixel localization, crucial for the task of accurate feature point tracking. We also demonstrate the effectiveness of our learning formulation in extensive feature point tracking experiments. Code and supplementary material are available at http://www.cvl.isy.liu.se/research/objrec/visualtracking/conttrack/index.html.Keywords
All Related Versions
- Version 1, 2016-08-12, ArXiv
- Version 2, 2016-08-29, ArXiv
- Published version: , 472.
This publication has 35 references indexed in Scilit:
- Fast Segmentation of Sparse 3D Point Trajectories Using Group Theoretical Invariants2015
- MEEM: Robust Tracking via Multiple Experts Using Entropy Minimization2014
- Transfer Learning Based Visual Tracking with Gaussian Processes Regression2014
- A Naturalistic Open Source Movie for Optical Flow Evaluation2012
- Exploiting the Circulant Structure of Tracking-by-Detection with Kernels2012
- P-N learning: Bootstrapping binary classifiers by structural constraints2010
- Visual object tracking using adaptive correlation filters2010
- Lucas-Kanade 20 Years On: A Unifying FrameworkInternational Journal of Computer Vision, 2004
- Improving Feature Tracking with Robust StatisticsPattern Analysis and Applications, 1999
- Good features to track1994