NutritionVerse-Synth（NV-Synth）是由滑铁卢大学创建的大型合成食品图像数据集，包含84,984张高分辨率的2D食品图像，源自7,082个程序生成的3D场景。该数据集通过动态和程序化两种方法生成，确保了食品场景的多样性和真实性。每个场景从12个不同视角捕捉，并附有详尽的营养信息和多模态标注，如RGB、深度、语义分割等。NV-Synth旨在推动计算机视觉技术在自动化饮食评估中

arXiv2023-12-11 更新220

Multilabel cuisine classification on US data

菜系分类

多标签分类

Multiple cuisine type and assign multiple cuisine types where applicable.

kaggle2022-09-24 更新140

Eitanli/cuisine_type

菜系分类

食谱数据

--- dataset_info: features: - name: id dtype: int64 - name: recipe dtype: string - name: cuisine_type dtype: string splits: - name: train num_bytes: 107643157 num_examp

Hugging Face2023-11-13 更新220

A catalog of Food Images Dataset

食品图像

机器学习

To build machine learning models for several tasks such as food recognition, food detection, food logging, calorie estimation, food perception, ingredient discovery, recipe retrieval, etc., there is a

DataCite Commons2023-10-05 更新150

Cusines dataset

资源简介：

相关数据集