Large Scale Visual Food Recognition

Weiqing Min1     Zhiling Wang2     Yuxin Liu1     Mengjiang Luo1     Liping Kang2     Xiaoming Wei2     Xiaolin Wei2     Shuqiang Jiang1    

1Institute of Computing Technology, Chinese Academy of Sciences; 2Meituan Group


This is the supplementary material for the paper "Large Scale Visual Food Recognition". The web intends to provide more details about our dataset, including the information of super-class distribution, the full category list, and samples for each category. More images from some categories are also provided.

The statistics of super-class distribution

We show the distribution of super-classes in Food2K in Fig.1. In Food2K, there are 12 super-classes and 26 sub-classes, which covering common types of existing food categories.

Fig.1: Distribution of super-classes in Food2K.

The full list of categories and samples of each category

The full list of categories and the number of images for each category here.
Samples of each category in Food2K here.

More image samples from some categories

     Margherita pizza

     Black pepper steak


     Mushroom pot

     Chocolate toast

     Leek box

     Crab seed sushi

     Wonton noodles

     Roasted eggplant with garlic

     Griddle chicken


This dataset can be obtained by sending a request email to us. Specifically, the researchers interested in it should download and fill up this Agreement Form and send it back to us (; Email title: Food2K dataset request). We will then send you the download instructions.

Code & Trained Models

You can download codes and trained models from here.

Related Paper

Weiqing Min, Zhiling Wang, Yuxin Liu, Mengjiang Luo, Liping Kang, Xiaoming Wei, Xiaolin Wei, Shuqiang Jiang*. Large Scale Visual Food Recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), DOI:10.1109/TPAMI.2023.3237871, 2023. [link].