The Met Dataset
Met

What is the Met Dataset?


The Met Dataset is a large-scale dataset for Instance-Level Recognition (ILR) in the artwork domain.

Evaluation


Recognition performance on the test set of the Met Dataset is measured with two standard metrics, namely average classification accuracy (ACC), and Global Average Precision (GAP).

Downloads

Dataset


The images of the dataset and the ground-truth files can be downloaded from the links below. All images have been resized so that their largest side is 500 pixels.

Descriptors


Extracted descriptors (multiscale) for different training methods can be downloaded from the links below. They can be used directly with the given code for kNN classification.

Models


Trained models on the Met training set for different training methods can be downloaded from the links below. They can be used directly to extract descriptors with the given code for descriptor extraction.

Code


Example code demonstrating how to use the dataset, perform the evaluation and reproduce experiments with the baseline approaches is available here.

Related Publication


The Met Dataset: Instance-level Recognition for Artworks (pdf)
N.A. Ypsilantis, N. Garcia, G. Han, S. Ibrahimi, N. van Noord, G. Tolias
Accepted at NeurIPS 2021 Track on Datasets and Benchmarks.

Feedback

Any feedback is very welcome. Please send it to: nikosips98@gmail.com