Studying aesthetics in photographic images using a computational approach
The aesthetic quality of a picture is automatically inferred using visual content as a machine learning problem using, for example, a peer-rated, on-line photo sharing Website as data source. Certain visual features of images are extracted based on the intuition that they can discriminate between aesthetically pleasing and displeasing images. A one-dimensional support vector machine is used to identify features that have noticeable correlation with the community-based aesthetics ratings. Automated classifiers are constructed using the support vector machines and classification trees, with a simple feature selection heuristic being applied to eliminate irrelevant features. Linear regression on polynomial terms of the features is also applied to infer numerical aesthetics ratings.
1. A computer-based method of inferring and utilizing aesthetic quality of photographs and other images, comprising the steps of:
receiving a plurality of digitized images along with aesthetic-based ratings of the images;
performing one or more software operations on the digitized images to automatically extract a plurality of visual features representative of each image;
receiving an image without an aesthetic-based rating;
automatically extracting a plurality of visual features representative of the received image;
computing a familiarity measure for the received image by correlating the visual features extracted from the received image to the visual features extracted from the other images;
determining an aesthetic-based rating for the received image on the basis of the familiarity measure, wherein a lower familiarity is indicative of originality and a higher rating; and
using one or more statistical methods to correlate the extracted visual features and the aesthetic-based ratings to classify the images on the basis of aesthetic value, rate the images on a scale relating to aesthetics value, or select/eliminate an image based upon aesthetic quality.
2. The method of claim 1 , wherein the correlation is performed using a support vector machine.
3. The method of claim 1 , further including the step of converting each image into the HSV color space to produce two-dimensional matrices IH, IS and IV for feature extraction.
4. The method of claim 1 , further including the step of automatically segmenting each image to identify objects in the image.
5. The method of claim 1 , wherein one of the plurality of features is color distribution.
6. The method of claim 1 , further including the step of performing a linear regression on one or more of the features to infer the aesthetic-based rating for the received image.
7. The method of claim 1 , further including the use of a naïve Bayes classifier to classify the image based on aesthetic-based rating.
8. The method of claim 1 , wherein linear regression and naïve Bayes classification methods are used in conjunction to select high-quality images or eliminate low-quality images from an image collection.
9. The method of claim 1 , wherein the step of extracting a plurality of visual features representative of each image includes the employment of a content-based image retrieval algorithm.