RecipeGen: A Benchmark for Real-World Recipe Image Generation

2025-03-07

AI Summary

Recipe image generation in food computing suffers from a lack of real-world, multimodal aligned data—particularly triple-aligned annotations linking recipe goals, stepwise instructions, and corresponding images. Method: This paper introduces Recipe3D, the first benchmark for goal-step-image triadic recipe generation. Curated via expert annotation and multi-source culinary platform scraping, it covers diverse ingredients, multi-step cooking procedures, varied cooking styles, and broad food categories, enabling fine-grained spatiotemporal alignment across goals, procedural steps, and images. Contribution/Results: We propose the first end-to-end verifiable evaluation protocol supporting cross-modal understanding and generation modeling. The fully open-sourced dataset (GitHub) establishes the first standardized benchmark for recipe generation, multi-step visual reasoning, and food AI—significantly advancing research and applications in culinary intelligence.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Recipe image generation is an important challenge in food computing, with applications from culinary education to interactive recipe platforms. However, there is currently no real-world dataset that comprehensively connects recipe goals, sequential steps, and corresponding images. To address this, we introduce RecipeGen, the first real-world goal-step-image benchmark for recipe generation, featuring diverse ingredients, varied recipe steps, multiple cooking styles, and a broad collection of food categories. Data is in https://github.com/zhangdaxia22/RecipeGen.

Problem

• Lack of real-world dataset for recipe image generation
• Need to connect recipe goals, steps, and images
• Introduction of RecipeGen benchmark for diverse recipe generation

Innovation

• First real-world goal-step-image benchmark
• Diverse ingredients and cooking styles
• Comprehensive dataset for recipe generation