Multiple People Tracking by Lifted Multicut and Person Re-identification
Citations
2,016 citations
Cites background from "Multiple People Tracking by Lifted ..."
...Visual object tracking is a basic building block in various tasks of computer vision, such as automatic driving [19] and video surveillance [32]....
[...]
737 citations
Cites background from "Multiple People Tracking by Lifted ..."
...A graph-based formulation to link person hypotheses is proposed for multi-person tracking [203], where the holistic features of the full human body and body pose layout are combined as the representation for each person....
[...]
657 citations
Cites background or methods from "Multiple People Tracking by Lifted ..."
...We believe that there is an exciting avenue for future work in combining local trackers (such as our work) with stronger offline long-range models (such as SORT [2], LMP [41], and other ReID-based trackers [50, 52])....
[...]
...With the advent of high-performing object detection models [9, 31], a powerful alternative emerged: tracking-by-detection (or more precisely, tracking-after-detection) [2, 41, 50]....
[...]
...We follow prior works [32, 39, 41, 52, 55] to pretrain on external data....
[...]
...[41] leverage person-reidentification features and human pose features....
[...]
...Most modern trackers [2, 7, 23, 33, 35, 41, 47, 50, 58] follow the tracking-by-detection paradigm....
[...]
507 citations
Cites background from "Multiple People Tracking by Lifted ..."
...(Tang et al., 2017) leverage body pose features to enhance the appearance features....
[...]
...Tang et al. (Tang et al., 2019) detect object tubes in videos which aims to enhance classification scores in challenging frames based on their neighboring frames....
[...]
...There are also some works (Bae and Yoon, 2014; Tang et al., 2017; Sadeghian et al., 2017; Chen et al., 2018a; Xu et al., 2019) focusing on enhancing appearance features....
[...]
...Similar ideas have also been explored in (Han et al., 2016; Kang et al., 2016, 2017; Tang et al., 2019; Pang et al., 2020)....
[...]
...Tang et al. (Tang et al., 2017) leverage body pose features to enhance the appearance features....
[...]
503 citations
References
49,914 citations
12,531 citations
10,161 citations
"Multiple People Tracking by Lifted ..." refers methods in this paper
...Our implementation is based on the Caffe deep learning framework [11]....
[...]
5,308 citations
"Multiple People Tracking by Lifted ..." refers methods in this paper
...Following a common practice in face recognition/verfication literature [22], we use our ID-Net as initialization for learning the SiameseNet, StackNet and StackNetPose, which makes the training faster and produces better results....
[...]
3,564 citations
"Multiple People Tracking by Lifted ..." refers methods in this paper
...We also collect person identity examples from the CUHK03 [19], Market-1501 [37] datasets that are captured by 6 surveillance cameras....
[...]