Beyond Correlation Filters: Learning Continuous Convolution Operators for Visual Tracking

By Martin Danelljan, Andreas Robinson, Fahad Shahbaz Khan, Michael Felsberg. European Conference on Computer Vision, 2016

1,763 citations recorded by the source corpus.

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 b ...

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/5fd235751a9a3e79cfd7599f5e8b4ce7d7baf801

How researchers see this paper

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

In Computer Vision and Pattern Recognition

Citations in the corpus

Siamese Cascaded Region Proposal Networks for Real-Time Visual Tracking (2018, methodology): “Inspired by the successes in image classi-ﬁcation [21, 17], deep convolutional neural network (CNN) has been introduced into visual tracking and demonstrated excellent performances [49, 48, 34, 7, 12, 31, 8, 42].”
A Twofold Siamese Network for Real-Time Object Tracking (2018, background): “Some trackers [7, 26, 30, 9, 6] integrate deep features into conventional tracking approaches and benefit from the expressive power of CNN features.”
TrackingNet: A Large-Scale Dataset and Benchmark for Object Tracking in the Wild (2018, background): “In addition, many works enhance the original formulation by adapting the regression target [29], adding context [30, 31], spatially regularizing the learned ﬁlters and learning continuous ﬁlters [32].”
ATOM: Accurate Tracking by Overlap Maximization (2018, methodology): “CCOT [7] and UPDT [3], both based on correlation filters, achieve AUC scores of 49.”
RGB-T Object Tracking: Benchmark and Baseline (2018, methodology): “To identify the importance of thermal data, we select several advanced RGB trackers for evaluations, including ECO [49], C-COT [48], CFnet [39], DSST [50], SOWP [11], CSRDCF [53], SRDCF [52] and SAMF [51].”
Real-Time MDNet (2018, methodology): “For comparison, we employ several state-of-the-art trackers including ECO [3], MDNet [1], MDNet+IEL, SRDCF [31], C-COT [4], and top performing real-time trackers, ECO-HC [3], BACF [11], PTAV [12], CFNet [21], SiamFC [20] and DSST [32].”
Sim4CV: A Photo-Realistic Simulator for Computer Vision Applications (2017, neutral): “2014), C-COT (Danelljan et al. 2016) andMOSSECA(Mueller et al.”
Empowering Things With Intelligence: A Survey of the Progress, Challenges, and Opportunities in Artificial Intelligence of Things (2020, methodology): “Later, different deep learning methods have been proposed to improve the classical methods by learning multi-resolution deep features [72], end-to-end representation learning [73], and leveraging siamese networks [74].”
Learning Adaptive Discriminative Correlation Filters via Temporal Consistency Preserving Spatial Feature Selection for Robust Visual Object Tracking (2018, background): “proposed tracking method with 13 state-of-the-art trackers, including Staple [42], SRDCF [13], KCF [8], CSRDCF [14], STAPLE CA [82], BACF [15], C-COT [58], ECO [59], ACFN∗ [83], CREST∗ [37], SiamFC∗ [35], CFNet∗ [36], MCPF∗ [84], C-COT∗ and ECO∗.”
Meta-Tracker: Fast and Robust Online Adaptation for Visual Object Trackers (2018, background): “learn) studies [15,16,17,18,19,20], we seek to learn how to obtain the target model. The key idea is to train the target model in a way that generalizes well over future frames. In all previous works [1,2,5,6,7,8,9,10], the target model is trained to minimize a loss function on the current frame. Ev”
Multi-cue Correlation Filters for Robust Visual Tracking (2018, methodology): “In Figure 1, SRDCF [12], DeepSRDCF [11] and C-COT [13] adopt different types of features, but none of them is able to handle various challenging factors, even for the CCOT algorithm with multiple features as well as the novel continuous convolution operators.”
GOT-10k: A Large High-Diversity Benchmark for Generic Object Tracking in the Wild (2018, methodology): “For example, ECO improves CCOT in several ways and it achieves a much better performance than CCOT on OTB2015, but on GOT-10k it’s AO score is worse than CCOT.”
SiamCAR: Siamese Fully Convolutional Classification and Regression for Visual Tracking (2019, neutral): “62 CPU Matlab CCOT [24] 0.”
Distractor-aware Siamese Networks for Visual Object Tracking (2018, methodology): “We evaluate the proposed algorithms with numerous fast and state-of-the-art trackers including SiamFC [2], CFNet [33], Staple [1], CSRDCF [19], BACF [11], ECO-HC [3], CREST [29], MDNet [23], CCOT [5], ECO [3], and the baseline tracker SiamRPN [16].”
Probabilistic Regression for Visual Tracking (2020, methodology): “This class of methods include the widely popular Discriminative Correlation Filter (DCF) approaches [5, 10, 12, 22], most of the more recent Siamese networks [2, 17, 45, 51] and other earlier approaches [50].”