Human Perception of Visual Information by Unknown

Author:Unknown
Language: eng
Format: epub
ISBN: 9783030814656
Publisher: Springer International Publishing

3.5 Additional Labels and Attributes

In addition to aesthetic scores, datasets can offer additional labels to enable multi-task applications (Kao et al. 2017b), or provide contextual information for aesthetic prediction. Typical additional labels include the semantic class of the picture, generally categorized based on the content, e.g., nature, portraits, buildings, etc. In some cases, the aesthetic data is complemented by textual annotations and comments crawled from the web or collected during the experiments. The text information has been used to provide aesthetic explanations, leveraging natural language processing architectures (Wang et al. 2019). Perceptual attributes directly contribute to aesthetic judgments, and some datasets focus on measuring them, although not in an aesthetic context. It is the case, for example, for colorfulness (Zerman et al. 2019) or dynamic range (Hulusic et al. 2016). Other datasets provide additional attributes such as the emotional response, which are not directly related to aesthetics, but can participate in image preference formation (Yu et al. 2019). Finally, aesthetic scores can be augmented with unique identifiers of voters, to facilitate personalized aesthetics applications.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.