A Review of YOLO Object Detection Model in Forensic Evidence Analysis

A Review of YOLO Object Detection Model in Forensic Evidence Analysis

Pooja Kolhatkar1, Dipak Padule2, Pratima Salunkhe3

1Department of Forensic Science YCIS, Satara.

2Software Engineer, SPPU, Pune, IN

3Professor, Department of Forensic Science YCIS, Satara.

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract - The rate of crimes is an ever-increasing problem in our society. A huge number of cases are still pending and need to be solved. Collecting and analysing evidence is becoming a more important complex task.  And it is becoming brain storming thing for officers.

With the availability of enormous data and the need for computer vision systems, research on object detection models/algorithms has become crucial over the last few decades. With various CNN architectures available, the state-of-the-art You Look Only Once (YOLO) model is very popular for many reasons mainly fast detection even on low-end devices.

Followed by introduction and background this paper reviews the innovative and descriptive approach YOLO takes at object detection and how it is helpful in Forensic Evidence Detection and Analysis.

Key Words:  YOLO, Forensic Evidence, Object Detection, Crime Scene