We are very happy to announce that our paper, entitled “From Impressionism to Expressionism: Automatically Identifying Van Gogh’s Paintings” has been accepted at the 23rd IEEE International Conference on Image Processing (ICIP) for ORAL presentation. ICIP is the world’s largest and most comprehensive technical conference focused on image and video processing and computer vision.
The accepted paper deals with the automatic identification of Vincent van Gogh’s paintings using a Convolutional Neural Network (CNN). Given a set of paintings from a painter of interest, and a number of images from other painters considering some sampling rules, the proposed method’s pipeline consists of dividing each image into smaller patches, extracting their discriminative visual features using CNN, training a patch classifier, and then using the relevant classification scores for a final aggregated response. The authors find out that using the patch with highest confidence score leads to the best result, outperforming alternative schemes.
The slides presented at ICIP are available here.
The adopted dataset is available here.