Prompt-Based Crime Scene Generation Using a Pre-Trained Diffusion Model
Abstract
In forensic investigations crime scene visualization plays an important and core task. It has also been used extensively in criminal justice education, and law-enforcement training. However, obtaining diverse and realistic crime scene imagery is often challenging due to privacy concerns, legal restrictions, and the limited availability of publicly accessible datasets. Diffusion based generative models have significantly advanced recently and show power and remarkable capabilities in synthesizing realistic images from natural language descriptions. Although the model has been explored for creative content generation, virtual scene synthesis, domain-specific image production and text-to-image frameworks for generating contextual visual content across healthcare, education, and simulation environments. However, the application of diffusion models to crime scene generation remains relatively underexplored. In this research we investigate the feasibility of prompt-based crime scene generation using a pre-trained Stable Diffusion model. Four representative crime scene categories, namely burglary, accident, vandalism, and arson, were defined through carefully designed textual prompts. The diffusion model generated multiple synthetic crime scene images for each category. To assess semantic consistency between prompts and generated images we used CLIP-based similarity analysis. Quantitative evaluation revealed average CLIP similarity scores of 0.307 for arson scenes, 0.305 for…
Read the paper
Related studies
- Learning from First-Generation Qualitative Approaches in the IS Discipline: An Evolutionary View and Some Implications for Authors and Evaluators (PART 1/2)
- Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models
- Prosthesis: New Technological Opportunities and Innovative Biomedical Devices
- The history and future of neural modeling for cochlear implants
- Ryanodine receptors: waking up from refractoriness
- Novel Transgenic Rabbit Model Sheds Light on the Puzzling Role of Matrix Metalloproteinase-12 in Atherosclerosis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.