Table 3 Qualitative comparison of discriminative trackers-based human tracking within a camera
Item No. | Used discriminative trackers | Speed | Occlusion | Scale change | Shape deformation |
|---|---|---|---|---|---|
1 | JLF-based JPDAF (Rasmussen et al. [18]) | Low | √ | √ | √ |
2 | Sample-based JPDAF (Schulz et al. [19]) | Low | √ | √ | √ |
3 | Clustering-based JPDAF (Naqvi et al. [20]) | Low | √ | √ | √ |
4 | JPDAF revisited (Rezatofighi et al. [21]) | Moderate | √ | √ | √ |
5 | Reliability measure-driven MHT (Zúñiga et al. [22]) | High | √ | √ | √ |
6 | MHT revisited (Kim et al. [23]) | Moderate | √ | √ | √ |
7 | Multiple association-based MHT (Joo et al. [24]) | High | √ | √ | √ |
8 | Hierarchical MHT (Zulkifley et al. [25]) | Low | √ | √ | √ |
9 | EOM-based FNF (Zhang et al. [26]) | High | √ | √ | √ |
10 | Greedy algorithms-based FNF (Pirsiavash et al. [27]) | High | √ | √ | √ |
11 | Lagrangian relaxation-based FNF (Butt et al. [28]) | High | √ | √ | √ |
12 | Multi-way data association-based FNF (Wu et al. [29]) | Low | √ | √ | √ |