Online Object Tracking: A Benchmark 论文笔记(转)
2023-09-14 08:57:33 时间
转自:http://blog.csdn.net/lanbing510/article/details/40411877
有博主翻译了这篇论文:http://blog.csdn.net/roamer_nuptgczx/article/details/45790415
Factors that affect the performance of a tracing algorithm
1 Illumination variation
2 Occlusion
3 Background clutters
Main modules for object tracking
1 Target representation scheme
2 Search mechanism
3 Model update
Evaluation Methodology
1 Precison plot:
The percentage of frames whose estimated location is within the given threshold distance of the ground truth.
x coordinate: threshold
2 Success plot:
The ratios of successful frames at the thresholds varied from 0 to 1
x coordinate: threshold
3 Robustness Evaluation
A OPE: one-pass evaluation
B TRE temporal robustness evaluation
C SRE spatial robustness evaluation
Overall Performance
详见论文
1 TLD performs well in long sequences with a redetection module
2 Struck only estimates the location of target and does not handle scale variation
3 Sparse representations are effectivemodels to account for appearance change (e.g., occlusion).
4 Local sparse representations are more effective than the ones with holistic sparse
templates.
templates.
5 It indicates the alignmentpooling technique adopted by ASLA is more robust to misalignments and background clutters.
6 When an object moves fast, dense sampling based trackers (e.g., Struck, TLD and CXT) perform much better than others
7 On the OCC subset, the Struck, SCM, TLD, LSK and ASLA methods outperform others. The results suggest that structured learning and local sparse representations are effective in dealing with occlusions.
8 On the SV subset,ASLA, SCM and Struck perform best. The results show that
trackers with affine motion models (e.g., ASLA and SCM) often handle scale variation better than others that are designed to account for only translational motion with a few exceptions such as Struck
trackers with affine motion models (e.g., ASLA and SCM) often handle scale variation better than others that are designed to account for only translational motion with a few exceptions such as Struck
9 The performance of TLD, CXT, DFT and LOT decreases with the increase of
initialization scale. This indicates these trackers are more sensitive to background clutters.
initialization scale. This indicates these trackers are more sensitive to background clutters.
10 On the other hand, some trackers perform well or even better when the initial bounding box is enlarged, such as Struck, OAB, SemiT, and BSBT. This indicates that the Haar-like features are somewhat robust to background clutters due to the summation operations when computing features. Overall, Struck is less sensitive to scale variation than other well-performing methods.
11 Some trackers perform better when the scale factor is smaller, such as L1APG, MTT, LOT and CPF
//补充
Concluding Remarks
1.background information is critical for effective tracking.
2.local models are important for tracking
3.motion model or dynamic model is crucial for object tracking, especially when the motion of target
is large or abrupt
is large or abrupt
Good location prediction based on the dynamic model could reduce the search range and thus improve the tracking efficiency and robustness.
Dataset
对应网站
一篇教程:http://blog.csdn.net/carrierlxksuper/article/details/47054231
相关文章
- 耗时半年的论文,被审稿人一句没有novelty打发了?
- 2017美赛A题论文阅读笔记
- 论文写作清单
- CVPR2022论文速递(2022.6.14)!共9篇
- CVPR2022论文速递(2022.6.22)!共16篇!
- 谷歌Borg论文阅读笔记(一)—— 集群操作系统
- 谷歌Borg论文阅读笔记(二)—— 任务混部和资源隔离
- 论文笔记之STN_论文笔记软件
- 高斯过程相关研究的新进展的8篇论文推荐(统计 +人工智能)
- 【论文笔记】2019-ACL-Dialog State Tracking: A Neural Reading Comprehension Approach
- 【论文笔记】Multi-Domain Dialogue State Tracking based on State Graph
- 【论文笔记】Jointly Optimizing State Operation Prediction and Value Generation for Dialogue State Tracking
- ECCV2022|自校准高动态范围辐射场!论文速递2022.10.4!
- [AAAI 2022 | 论文简读] SpreadGNN: 分子数据上图神经网络的分散式多任务联邦学习
- 2022年最有开创性的10篇AI论文总结
- 论文笔记:《Bringing Balance to the Force Dynamic Analysis of the Android Application Framework》
- 《MIMIC-IV, a freely accessible electronic health record dataset》论文笔记
- 【论文笔记】Multi-Attention-Based Scalable Dialog State Tracking
- WWW-2021最佳论文亚军花落天津大学:基于属性补全的异质图神经网络新架构
- 引入N-gram改进Transformer架构,ACL匿名论文超越Primer等基准
- 【论文笔记】A Sequence-to-Sequence Approach to Dialogue State Tracking
- 在一篇由GPT-3生成的论文中,ChatGPT重现了图灵测试的原始论文
- 点石成金!AI成功挖掘论文“宝藏”
- IROS多项奖项公布!中国队伍称雄抓取竞赛,华人学生获最佳学生论文奖 | IROS 2017
- ICRA的5篇最佳自动化入围论文都有啥看点?| ICRA 2017