Gen-AI Police Sketches with Stable Diffusion

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Fidalgo, Nicholas, Contreras, Aaron, Harvey, Katherine, Ni, Johnny
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913959166083072
author Fidalgo, Nicholas
Contreras, Aaron
Harvey, Katherine
Ni, Johnny
author_facet Fidalgo, Nicholas
Contreras, Aaron
Harvey, Katherine
Ni, Johnny
contents This project investigates the use of multimodal AI-driven approaches to automate and enhance suspect sketching. Three pipelines were developed and evaluated: (1) baseline image-to-image Stable Diffusion model, (2) same model integrated with a pre-trained CLIP model for text-image alignment, and (3) novel approach incorporating LoRA fine-tuning of the CLIP model, applied to self-attention and cross-attention layers, and integrated with Stable Diffusion. An ablation study confirmed that fine-tuning both self- and cross-attention layers yielded the best alignment between text descriptions and sketches. Performance testing revealed that Model 1 achieved the highest structural similarity (SSIM) of 0.72 and a peak signal-to-noise ratio (PSNR) of 25 dB, outperforming Model 2 and Model 3. Iterative refinement enhanced perceptual similarity (LPIPS), with Model 3 showing improvement over Model 2 but still trailing Model 1. Qualitatively, sketches generated by Model 1 demonstrated the clearest facial features, highlighting its robustness as a baseline despite its simplicity.
format Preprint
id arxiv_https___arxiv_org_abs_2507_18667
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Gen-AI Police Sketches with Stable Diffusion
Fidalgo, Nicholas
Contreras, Aaron
Harvey, Katherine
Ni, Johnny
Computer Vision and Pattern Recognition
Artificial Intelligence
This project investigates the use of multimodal AI-driven approaches to automate and enhance suspect sketching. Three pipelines were developed and evaluated: (1) baseline image-to-image Stable Diffusion model, (2) same model integrated with a pre-trained CLIP model for text-image alignment, and (3) novel approach incorporating LoRA fine-tuning of the CLIP model, applied to self-attention and cross-attention layers, and integrated with Stable Diffusion. An ablation study confirmed that fine-tuning both self- and cross-attention layers yielded the best alignment between text descriptions and sketches. Performance testing revealed that Model 1 achieved the highest structural similarity (SSIM) of 0.72 and a peak signal-to-noise ratio (PSNR) of 25 dB, outperforming Model 2 and Model 3. Iterative refinement enhanced perceptual similarity (LPIPS), with Model 3 showing improvement over Model 2 but still trailing Model 1. Qualitatively, sketches generated by Model 1 demonstrated the clearest facial features, highlighting its robustness as a baseline despite its simplicity.
title Gen-AI Police Sketches with Stable Diffusion
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2507.18667