Case Study

Criminology images analysis

Through Computer vision methodology, our client wants to locate possible DNA areas on images coming from crime scenes to accelerate the process of analysis.​

Contact us for more detail

IN BRIEF

KEY INFORMATION

DNA LOCATION

Bounding box using deep learning

AUTO-LABEL

With an intelligent use of model’s probabilities​

GLOBAL-LEARNING

Generalization to items never encountered during the training phase​

Challenges

  • Generalize on objects the algorithm doesn’t know thanks to a future-proof model
  • Consider the laboratory confidence scores for heatmaps visualization
  • Create an autonomous solution which could be deployed on an isolated machine

Solution

  • Creation of an automatized labelling solution for possible DNA areas on images
  • Rectangles of localization & probabilistic heatmaps for results.
  • Time saving from the photo shooting to the sample

Gains

  • Pictures analyzed: 2000​
  • History of all analysis stored through time and easy to browse​
  • Ability to analyze any kind of objects found on crime scene​

Testimonial

“This solution has completely changed our way of working and we
no longer waste time”​