Skip to main content

Table 2 Anonymous video analytics algorithms for localization, age, and gender estimation

From: Benchmark for anonymous video analytics

Alias Ref. Name Person part Localization Age Gender Implementation Count performance Speed
     Detection Tracking    GPU CPU Instantaneous Cumulative  
           People OTS   
A1 [14] RetinaFace Face \(\checkmark\)     \(\checkmark\) \(\checkmark\) + +++ + +++
A2 [15] MTCNN Face \(\checkmark\)     \(\checkmark\) \(\checkmark\) + +++ + +++
A3 [16] DeepSORT Person   \(\checkmark\)    \(\checkmark\) \(\checkmark\) +++ + +++ ++
A4 [17] TRMOT Person   \(\checkmark\)    \(\checkmark\)   +++ + +++ ++
A5 [18] FaceLib Face \(\checkmark\) \(\checkmark\) \(\checkmark\) \(\checkmark\) +++
A6 [19] DEX Face   \(\checkmark\) \(\checkmark\) \(\checkmark\) \(\checkmark\) +++
C1 Commercial-1 Face   \(\checkmark\) \(\checkmark\) \(\checkmark\)   \(\checkmark\) + +++ +++ ++
C2 Commercial-2 Face   \(\checkmark\) \(\checkmark\) \(\checkmark\)   \(\checkmark\) +++ +++ +++ ++
  1. Algorithms count performance and speed are assessed, on a scale from 1 to 3 (the higher, the better) and regardless of the system used, based on generic algorithmic concepts (e.g., tracking outperforms detection for cumulative counts) and not on experimental results. OTS: opportunity to see. A4 was not compatible with OpenVINO optimization at submission time, thus we run A4 on CPU without optimization