English
Related papers

Related papers: Ocean: Object-aware Anchor-free Tracking

200 papers

Visual object tracking is an important application of computer vision. Recently, Siamese based trackers have achieved good accuracy. However, most of Siamese based trackers are not efficient, as they exhaustively search potential object…

Computer Vision and Pattern Recognition · Computer Science 2020-06-16 Shengyun Peng , Yunxuan Yu , Kun Wang , Lei He

The current Siamese network based on region proposal network (RPN) has attracted great attention in visual tracking due to its excellent accuracy and high efficiency. However, the design of the RPN involves the selection of the number,…

Computer Vision and Pattern Recognition · Computer Science 2021-04-16 Kai Yang , Zhenyu He , Wenjie Pei , Zikun Zhou , Xin Li , Di Yuan , Haijun Zhang

Trackers based on Siamese network have shown tremendous success, because of their balance between accuracy and speed. Nevertheless, with tracking scenarios becoming more and more sophisticated, most existing Siamese-based approaches ignore…

Computer Vision and Pattern Recognition · Computer Science 2020-08-11 Zhongzhou Zhang , Lei Zhang

In this paper, an online adaptive model-free tracker is proposed to track single objects in video sequences to deal with real-world tracking challenges like low-resolution, object deformation, occlusion and motion blur. The novelty lies in…

Computer Vision and Pattern Recognition · Computer Science 2017-12-12 Tanushri Chakravorty , Guillaume-Alexandre Bilodeau , Eric Granger

In the domain of visual tracking, most deep learning-based trackers highlight the accuracy but casting aside efficiency. Therefore, their real-world deployment on mobile platforms like the unmanned aerial vehicle (UAV) is impeded. In this…

Computer Vision and Pattern Recognition · Computer Science 2021-08-02 Changhong Fu , Ziang Cao , Yiming Li , Junjie Ye , Chen Feng

We present ObjectBox, a novel single-stage anchor-free and highly generalizable object detection approach. As opposed to both existing anchor-based and anchor-free detectors, which are more biased toward specific object scales in their…

Computer Vision and Pattern Recognition · Computer Science 2022-07-15 Mohsen Zand , Ali Etemad , Michael Greenspan

Recently, most siamese network based trackers locate targets via object classification and bounding-box regression. Generally, they select the bounding-box with maximum classification confidence as the final prediction. This strategy may…

Computer Vision and Pattern Recognition · Computer Science 2021-07-20 Jinlong Peng , Zhengkai Jiang , Yueyang Gu , Yang Wu , Yabiao Wang , Ying Tai , Chengjie Wang , Weiyao Lin

Siamese approaches have achieved promising performance in visual object tracking recently. The key to the success of Siamese trackers is to learn appearance-invariant feature embedding functions via pair-wise offline training on large-scale…

Computer Vision and Pattern Recognition · Computer Science 2020-05-29 Tianyang Xu , Zhen-Hua Feng , Xiao-Jun Wu , Josef Kittler

Accurate tracking is still a challenging task due to appearance variations, pose and view changes, and geometric deformations of target in videos. Recent anchor-free trackers provide an efficient regression mechanism but fail to produce…

Computer Vision and Pattern Recognition · Computer Science 2021-04-02 Yutao Cui , Cheng Jiang , Limin Wang , Gangshan Wu

In object detection, determining which anchors to assign as positive or negative samples, known as anchor assignment, has been revealed as a core procedure that can significantly affect a model's performance. In this paper we propose a…

Computer Vision and Pattern Recognition · Computer Science 2020-09-08 Kang Kim , Hee Seok Lee

In training object detector based on convolutional neural networks, selection of effective positive examples for training is an important factor. However, when training an anchor-based detectors with sparse annotations on an image, effort…

Computer Vision and Pattern Recognition · Computer Science 2020-06-23 Jihun Yoon , Seungbum Hong , Sanha Jeong , Min-Kook Choi

Object detection has been dominated by anchor-based detectors for several years. Recently, anchor-free detectors have become popular due to the proposal of FPN and Focal Loss. In this paper, we first point out that the essential difference…

Computer Vision and Pattern Recognition · Computer Science 2020-06-23 Shifeng Zhang , Cheng Chi , Yongqiang Yao , Zhen Lei , Stan Z. Li

In this paper, we propose a general approach to optimize anchor boxes for object detection. Nowadays, anchor boxes are widely adopted in state-of-the-art detection frameworks. However, these frameworks usually pre-define anchor box shapes…

Computer Vision and Pattern Recognition · Computer Science 2020-01-28 Yuanyi Zhong , Jianfeng Wang , Jian Peng , Lei Zhang

Modern CNN-based object detectors assign anchors for ground-truth objects under the restriction of object-anchor Intersection-over-Unit (IoU). In this study, we propose a learning-to-match approach to break IoU restriction, allowing objects…

Computer Vision and Pattern Recognition · Computer Science 2019-11-13 Xiaosong Zhang , Fang Wan , Chang Liu , Rongrong Ji , Qixiang Ye

Robustness and discrimination power are two fundamental requirements in visual object tracking. In most tracking paradigms, we find that the features extracted by the popular Siamese-like networks cannot fully discriminatively model the…

Computer Vision and Pattern Recognition · Computer Science 2022-03-04 Fei Xie , Chunyu Wang , Guangting Wang , Yue Cao , Wankou Yang , Wenjun Zeng

Existing deep trackers mainly use convolutional neural networks pre-trained for generic object recognition task for representations. Despite demonstrated successes for numerous vision tasks, the contributions of using pre-trained deep…

Computer Vision and Pattern Recognition · Computer Science 2019-04-04 Xin Li , Chao Ma , Baoyuan Wu , Zhenyu He , Ming-Hsuan Yang

Similarity matching is a core operation in Siamese trackers. Most Siamese trackers carry out similarity learning via cross correlation that originates from the image matching field. However, unlike 2-D image matching, the matching network…

Computer Vision and Pattern Recognition · Computer Science 2021-05-07 Jinpu Zhang , Yuehuan Wang

We propose a novel and flexible anchor mechanism named MetaAnchor for object detection frameworks. Unlike many previous detectors model anchors via a predefined manner, in MetaAnchor anchor functions could be dynamically generated from the…

Computer Vision and Pattern Recognition · Computer Science 2018-11-07 Tong Yang , Xiangyu Zhang , Zeming Li , Wenqiang Zhang , Jian Sun

Most of the existing trackers usually rely on either a multi-scale searching scheme or pre-defined anchor boxes to accurately estimate the scale and aspect ratio of a target. Unfortunately, they typically call for tedious and heuristic…

Computer Vision and Pattern Recognition · Computer Science 2020-04-23 Zedu Chen , Bineng Zhong , Guorong Li , Shengping Zhang , Rongrong Ji

Existing anchor-based and anchor-free object detectors in multi-stage or one-stage pipelines have achieved very promising detection performance. However, they still encounter the design difficulty in hand-crafted 2D anchor definition and…

Computer Vision and Pattern Recognition · Computer Science 2020-05-13 Geng Zhan , Dan Xu , Guo Lu , Wei Wu , Chunhua Shen , Wanli Ouyang
‹ Prev 1 2 3 10 Next ›