Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Criminator: An Easy-to-Use XR "Crime Animator" for Rapid Reconstruction and Analysis of Dynamic Crime Scenes

Vahid Pooryousef, Lonni Besançon, Maxime Cordeil, Chris Flight, Alastair M Ross AM, Richard Bassed, Tim Dwyer

TL;DR

Criminator addresses the need for accessible dynamic crime-scene reconstruction in XR by introducing a co-design-informed framework and a track-slot animation system with Gaussian-Splatting environments. The study demonstrates that both criminology-trained and lay participants can observe and author animated scenes, supporting uses in hypothesis testing, demonstrations, and training while highlighting ethical and legal considerations for admissibility and provenance. It integrates expert input to balance scientific rigor with usability, showing promise for democratizing judicial visualization and education. The work advances forensic visualization by combining usability, controllable realism, and explainability, with practical implications for investigative analysis and training in law enforcement and legal contexts.

Abstract

Law enforcement authorities are increasingly interested in 3D modelling for virtual crime scene reconstruction, enabling offline analysis without the cost and contamination risk of on-site investigation. Past work has demonstrated spatial relationships through static modelling but validating the sequence of events in dynamic scenarios is crucial for solving a case. Yet, animation tools are not well suited to crime scene reconstruction, and complex for non-experts in 3D modelling/animation. Through a co-design process with criminology experts, we designed "Criminator"-a methodological framework and XR tool that simplifies animation authoring. We evaluated this tool with participants trained in criminology (n=6) and untrained individuals (n=12). Both groups were able to successfully complete the character animation tasks and provided high usability ratings for observation tasks. Criminator has potential for hypothesis testing, demonstration, sense-making, and training. Challenges remain in how such a tool fits into the entire judicial process, with questions about including animations as evidence.

Criminator: An Easy-to-Use XR "Crime Animator" for Rapid Reconstruction and Analysis of Dynamic Crime Scenes

TL;DR

Criminator addresses the need for accessible dynamic crime-scene reconstruction in XR by introducing a co-design-informed framework and a track-slot animation system with Gaussian-Splatting environments. The study demonstrates that both criminology-trained and lay participants can observe and author animated scenes, supporting uses in hypothesis testing, demonstrations, and training while highlighting ethical and legal considerations for admissibility and provenance. It integrates expert input to balance scientific rigor with usability, showing promise for democratizing judicial visualization and education. The work advances forensic visualization by combining usability, controllable realism, and explainability, with practical implications for investigative analysis and training in law enforcement and legal contexts.

Abstract

Law enforcement authorities are increasingly interested in 3D modelling for virtual crime scene reconstruction, enabling offline analysis without the cost and contamination risk of on-site investigation. Past work has demonstrated spatial relationships through static modelling but validating the sequence of events in dynamic scenarios is crucial for solving a case. Yet, animation tools are not well suited to crime scene reconstruction, and complex for non-experts in 3D modelling/animation. Through a co-design process with criminology experts, we designed "Criminator"-a methodological framework and XR tool that simplifies animation authoring. We evaluated this tool with participants trained in criminology (n=6) and untrained individuals (n=12). Both groups were able to successfully complete the character animation tasks and provided high usability ratings for observation tasks. Criminator has potential for hypothesis testing, demonstration, sense-making, and training. Challenges remain in how such a tool fits into the entire judicial process, with questions about including animations as evidence.
Paper Structure (36 sections, 10 figures, 3 tables)

This paper contains 36 sections, 10 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Background
  3. Benefits and Concerns of Visual Fidelity-Levels of CSR
  4. CSR using XR
  5. Authoring Animation using XR
  6. Method
  7. Criminator System Design
  8. Experts' Insights
  9. Mechanisms
  10. Criminator Features
  11. User Interface
  12. Effects and Controllable Objects for CSR
  13. Animation Recorder and Player
  14. User Study
  15. Scenario
  16. ...and 21 more sections

Figures (10)

  • Figure 1: Our methodology for design, evaluation and analysis.
  • Figure 2: Overview of CSR in VR mechanism.
  • Figure 3: Animation Editor UI.
  • Figure 4: Effects (3 left) and props (right) panels. The Effects section includes: (left) a list of current effects attached to a slot, shown as "target prop:effect"; (middle) a list of all available effects that can be added to a slot; and (right) the parameters of the selected effect.
  • Figure 5: The overview of the user study's scenario.
  • ...and 5 more figures