Can we teach computers to understand art? Domain adaptation for enhancing deep networks capacity to de-abstract art

作者：

Highlights：

• We train from scratch a deep network to reach top performance in painting genre recognition.

• We identify the style abstraction as the main limitation while trying to improve.

• We test the potential for improvement by transfer from various domains.

• Normal photographs (no adaptation) are as beneficial as style transfer.

摘要

•We train from scratch a deep network to reach top performance in painting genre recognition.•We identify the style abstraction as the main limitation while trying to improve.•We test the potential for improvement by transfer from various domains.•Normal photographs (no adaptation) are as beneficial as style transfer.

论文关键词：Convolutional Neural Networks,Domain adaptation,Genre recognition,Painting analysis,Style transfer

论文评审过程：Received 11 December 2017, Accepted 21 June 2018, Available online 4 July 2018, Version of Record 20 July 2018.

论文官网地址：https://doi.org/10.1016/j.imavis.2018.06.009