Recipe Generation from Food Images Using Deep Learning

Abstract: The advancement of artificial intelligence in computer vision and natural language processing has enabled innovative applications such as automatic recipe generation from food images. This paper presents a deep learning- based system that analyzes food images to predict ingredients and generate step-by-step cooking instructions. The proposed system integrates Convolutional Neural Networks (CNNs) for extracting visual features and Natural Language Processing (NLP) models such as Long Short-Term Memory (LSTM) or Transformers for generating recipes. The system aims to assist users in identifying unknown dishes and preparing them efficiently. Experimental observations indicate that the model produces contextually relevant and grammatically correct recipes. This approach has potential applications in smart kitchens, food blogging, and diet planning systems.

Keywords: Food Image Recognition, Recipe Generation, Deep Learning, CNN, NLP, Image Captioning