The objective of this research is to provide a barrierfree artwork appreciation experience for people with visual impairment (PVI) who have limitations because of the lack of cognitive and sensory access to exhibitions, museums, etc. We applied artifi...
The objective of this research is to provide a barrierfree artwork appreciation experience for people with visual impairment (PVI) who have limitations because of the lack of cognitive and sensory access to exhibitions, museums, etc. We applied artificial intelligence (AI) to our previous works with visual appreciation solutions recommending poets according to the colors and objects. To extract the color and object elements in the poet, we used a natural language toolkit (NLTK) and in the visual art, we used convolutional neural networks (CNNs) to train the dataset of moon and church images. Unlike the researches of classification on natural images, despite the small size of the dataset and numerous variables in the visual arts such as techniques (oil, watercolor, etc.) and styles (impressionism, modernism, etc.), we attained successful results.