Pillow-1/Xsyn

Taming Generative Synthetic Data for X-ray Prohibited Item Detection

This repository contains the Xsyn model, a one-stage X-ray security image synthesis pipeline based on text-to-image generation. Proposed in the paper Taming Generative Synthetic Data for X-ray Prohibited Item Detection, Xsyn addresses data insufficiency for prohibited item detection by incorporating two effective strategies: Cross-Attention Refinement (CAR) for refining bounding box annotations and Background Occlusion Modeling (BOM) for enhancing imaging complexity. It aims to achieve high-quality X-ray security image synthesis without incurring additional labor-intensive foreground preparation.

Code repository: https://github.com/pILLOW-1/Xsyn/

Download Xsyn models

Checkpoints for different datasets are available. All models here are based on GLIGEN.

Dataset	Mode	Download
PIDray	text-grounded inpainting	HF Hub
OPIXray	text-grounded inpainting	HF Hub
HiXray	text-grounded inpainting	HF Hub

Inference

We provide one script to generate x-ray security images and construct their annotations. First download models and put them in --ckpt_path. Then run:

python gligen_inference.py

Details of some important args:

• --output_path: the path to save your generated x-ray security images
• --annotation_path: the path to save the refined annotation (stored in txt format)
• --vis_path: the path to save visualization compared with gt
• --ca_vis_path: the path to save cross-attention maps
• --image_path: the path to load images you want to inpaint
• --ckpt_path: the generation model checkpoint path
• --gligen_caption_pt: the file to prepare your training/test data in GLIGEN format
• --gen_method: set to 1 for Xsyn-M and 3 for Xsyn-A
• --refine_anno: set to True for CAR
• --latent_redist: set to True for BOM

After inference, you can use downstream_test.sh to test the performance of our synthetic data. Our downstream detection environment is mmdetection.