Modeling Image Classification with Deep Learning Algorithms
University of Bern Releases Unique Abstract Art Dataset for Image Classification
Researchers at the University of Bern in Switzerland have made a significant contribution to the field of image classification by releasing a dataset of abstract art. This dataset, credited to DELAUNAY, contains 11,503 images from 53 different artists, specifically designed for training image classification models.
The dataset is not geographically restricted and can be used for recognising images with unusual structures. It's important to note that no new images are added to the dataset. The dataset is for training purposes only, and it's specifically designed for abstract art.
While a direct link to the dataset may not be readily available, there are several ways to access it. One approach is to visit or contact the University of Bern's departments likely involved, such as the Institute of Computer Science or the Department of Art History, directly. Although contact emails of some professors and researchers at the University of Bern were found in relation to other research topics, none were specifically related to an abstract art dataset.
Another method is to check if there are any publications or project pages by University of Bern researchers that might reference such a dataset, as research datasets are often linked with published papers.
If the dataset is not publicly hosted by the University of Bern, exploring commonly used public datasets in abstract art image classification could be an alternative.
Since none of the retrieved documents mention or link to a University of Bern abstract art dataset specifically focused on image classification, if this dataset exists, it is likely to be accessed through official university research collaborations or specific project repositories not indexed in the current searched pages.
If you require assistance in finding alternative well-known abstract art image datasets for training classification models, I'd be happy to help further.
The dataset, credited to DELAUNAY and specifically designed for training image classification models, can be utilized for recognizing images with unusual structures, demonstrating its potential use in artificial-intelligence research. Researchers may also consider exploring other public datasets in abstract art image classification if the University of Bern dataset is not publicly hosted or if they encounter difficulties in accessing it.
