Deeper and Wider Siamese Networks for Real-Time Visual Tracking

By Zhipeng Zhang, Houwen Peng, Qiang Wang. Computer Vision and Pattern Recognition, 2019

954 citations recorded by the source corpus.

Siamese networks have drawn great attention in visual tracking because of their balanced accuracy and speed. However, the backbone networks used in Siamese trackers are relatively shallow, such as AlexNet, which does not fully take advantage of the capability of modern deep neural networks. In this paper, we investigate how to leverage deeper and wider convolutional neural networks to enhance tracking robustness and accuracy. We observe that direct replacement of backbones with existing powerful architectures, such as ResNet and Inception, does not bring improvements. The main reasons are that 1) large increases in the receptive field of neurons lead to reduced feature discriminability and localization precision; and 2) the network padding for convolutions induces a positional bias in learning. To address these issues, we propose new residual modules to eliminate the negative impact of padding, and further design new architectures using these modules with controlled receptive field size and network stride. The designed architectures are lightweight and guarantee real-time tracking speed when applied to SiamFC and SiamRPN. Experiments show that solely due to the proposed network architectures, our SiamFC+ and SiamRPN+ obtain up to 9.8%/6.3% (AUC), 23.3%/8.8% (EAO) and 24.4%/25.0% (EAO) relative improvements over the original versions on the OTB-15, VOT-16 and VOT-17 datasets, respectively.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/530c0e08362aceb66c226fa4379523820ef56ba3

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

The Seventh Visual Object Tracking VOT2019 Challenge Results (2019, methodology): “In RPN network, we further increase the depth of the backbone network in SiamDW [109], and replace it with a much deeper one, i.”
Siamese Box Adaptive Network for Visual Tracking (2020, methodology): “SA SIAM R [18] SiamCRF RT [18] SPM [39] SiamRPN++ [20] SiamMask [42] ARTCS [18] SiamDW ST [49] DCFST [18] DiMP [2] Ours”
Deep Learning for Visual Tracking: A Comprehensive Survey (2019, methodology): “vii) using deeper and wider deep network as the backbone to increase receptive field of neurons which is equivalent to capturing the structure of the target [56].”
LaSOT: A High-quality Large-scale Single Object Tracking Benchmark (2020, background): “, MDNet [68], TRACA [9], CFNet [76], SiamFC [6], StructSiam [92], DSiam [33], SINT [75], ATOM [15], DiMP [7], VITAL [74], SiamRPN++ [48], DaSiamRPN [95], SiamDW [93], C-RPN [27] and SiamMask [82], GlobalTrack [42]), correlation trackers with hand-crafted features (e.”
Graph Attention Tracking (2020, positive): “Both SiamRPN++ [15] and SiamDW [35] investigate to deepen neural networks to improve the tracking performance.”
Is First Person Vision Challenging for Object Tracking? (2020, positive): “We also considered deep siamese networks (SiamFC [8], GOTURN [36], DSLT [52], SiamRPN++ [48], SiamDW [91], UpdateNet [90], SiamFC++ [86], SiamBAN [15], Ocean [92]), trackingby-detection methods (MDNet [62], VITAL [73]), as well as trackers based on target segmentation representations (SiamMask [80],”
Deformable Siamese Attention Networks for Visual Object Tracking (2020, background): “Recently, SiamDW-RPN [43] and SiamRPN++ [24] were proposed, which allow the Siamese-based trackers to explore deeper networks, while Wang et al .”
Towards More Flexible and Accurate Object Tracking with Natural Language: Algorithms and Benchmark (2021, neutral): “339 023 SiamDW [61] CVPR-2019 Res22W 7 3 Local 150@GPU 0.”
Siam R-CNN: Visual Tracking by Re-Detection (2019, methodology): “Recent tracking approaches improve upon SiamRPN, making it distractor aware (DaSiamRPN [82]), adding a cascade (C-RPN [19]), producing masks (SiamMask [66]), using deeper architectures (SiamRPN+ [79] and SiamRPN++ [37]) and maintaining a set of diverse templates (THOR [58]).”