"FoodFake-30K: A Dataset for Real vs. AI-Generated Food Image Detection"
Abstract
"FoodFake-30K is a large-scale benchmark dataset for AI-generated food image detection, containing 30,000 images across 15 diverse food categories. The dataset includes 15,000 real food photographs sourced from Pexels, Unsplash, Pixabay, Reddit (r\/food), and Food-101, alongside 15,000 AI-generated images produced by two state-of-the-art generative models: Flux-2 Klein (9B-parameter Diffusion Transformer, FP8) and Z-Image Turbo (AuraFlow Diffusion Transformer, BF16).Real images underwent a rigorous preprocessing pipeline \u2014 CLIP-based semantic filtering (openai\/clip-vit-large-patch14), perceptual hash deduplication (pHash), and manual review. AI image prompts were generated using Qwen 2.5 Coder 14B with structured diversity slots (angle \u00d7 lighting \u00d7 scene \u00d7 lens) to maximize visual variety. Images are provided at 1024\u00d71024 resolution (27,400 images) and 512\u00d7512 resolution (2,600 Food-101 images). The 15 food categories span global cuisines: baklava, biryani, burger, cake\/pastry, croissant, dim sum, falafel, hummus, pizza, plov, ramen, salad, steak, sushi, and tacos \u2014 with 2,000 images per category (1,000 real + 500 Flux + 500 Z-Turbo). FoodFake-30K supports research in fake image detection, image forensics, food authenticity verification, deepfake detection benchmarking, and cross-category generalization studies. All metadata including generation prompts, seeds, and source URLs is provided under CC BY 4.0."
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.