Prediction of beauty and liking ratings for abstract and representational paintings using subjective and objective measures (original) (raw)
Related papers
The search for objective measures of aesthetic judgment: The case of memory traces
Empirical Studies of the Art, 2006
Verbal responses have frequently been used to measure aesthetic experience. They usually take the form of semantic judgments regarding specific aspects or dimensions of certain stimuli. The use of this kind of technique has produced a great amount of knowledge, but its combination with objective procedures can increase the validity and reliability of measurement. In this study, we set out to assess whether memory traces can serve as an objective control element for subjective aesthetic judgments. We analyzed the relation between aesthetic judgment and recognition of High Art and Popular Art visual stimuli by participants with and without formal art education. Results show that participants tended to give higher pleasantness and beauty ratings to those stimuli that have left a strong memory trace. Lower scores were awarded to stimuli they did not recognize well. However, originality and interest ratings did not follow the same trend. This disparity is discussed in relation to the dimensionality of aesthetic experience and the influence of formal art education on subjective measures of aesthetic experience.
INTEREST AND PLEASURE AS DIMENSIONS OF AESTHETIC RESPONSE
A study was conducted comparing cognitive and affective responses to a set of twelve paintings. forty subjects, including an equal number of naive and trained males and females, rated the paintings individually on a series of scales including: simple-complex, warm-cold, unemotional-emotional, not at all meaningful-meaningful, and familiar-unfamiliar. They also made comparative judgments of relative "interest" and "pleasure" between all possible pairs (66) of the paintings. "Interest" judgments were made under an objective and analytical task set, while "pleasing" judgments were made from a subjective and personal set. A multidimensional scaling analysis revealed two dimensions underlying the "pleasing" and "interest" judgments. These were identified by regressing the individual scale ratings against the dimensions. The "interest" dimensions, complexity/meaningfulness and familiarity, are comparable to motivational dimensions predicted by Bcrlyne: curiosity and variation. The "pleasing" dimensions, emotional arousal and aesthetic effectance, have been anticipated in the literature. Thematic content and date of the painting (modern versus pre-1850) could also be related to the dimensions. Individual differences in sensitivity to the "pleasing" dimensions were found primarily for naive and trained females.
2014 Sixth International Workshop on Quality of Multimedia Experience (QoMEX), 2014
A first step towards creating automatic measures of image aesthetic appeal is understanding its appreciation via subjective testing. Nevertheless, reliably setting up such tests appears to be challenging, as aesthetic appeal is proven to be influenced by a number of subjective factors. In this paper we investigate four scale types for aesthetic appeal rating and assess their ability to provide general quantification of aesthetic appeal, as well as repeatable judgment across experiments. We asked 24 users to assess a representative image set constructed to uniformly cover a wide range of aesthetic appeal. Our experiments show that the Absolute Category Rating (ACR) 5-point scale provides the most consistent ratings across participants, which are also repeatable across different experiments.
Subject Bias in Image Aesthetic Appeal Ratings
Data Science: Journal of Computing and Applied Informatics, 2017
Automatic prediction of image aesthetic appeal is an important part of multimedia and computer vision research, as it contributes to providing better content quality to users. Various features and learning methods have been proposed in the past to predict image aesthetic appeal more accurately. The effectiveness of these proposed methods often depend on the data used to train the predictor. Since aesthetic appeal is a subjective construct, factors that influence the subjectivity in aesthetic appeal data need to be understood and addressed. In this paper, we look into the subjectivity of aesthetic appeal data, and how it relates with image characteristics that are often used in aesthetic appeal prediction. We use subject bias and confidence interval to measure subjectivity, and check how they might be influenced by image content category and features.
Artificial intelligence has emerged as a powerful computational tool to create artworks. One application is Neural Style Transfer, which allows to transfer the style of one image, such as a painting, onto the content of another image, such as a photograph. In the present study, we ask how Neural Style Transfer affects objective image properties and how beholders perceive the novel (style-transferred) stimuli. In order to focus on the subjective perception of artistic style, we minimized the confounding effect of cognitive processing by eliminating all representational content from the input images. To this aim, we transferred the styles of 25 diverse abstract paintings onto 150 colored random-phase patterns with six different Fourier spectral slopes. This procedure resulted in 150 style-transferred stimuli. We then computed eight statistical image properties (complexity, self-similarity, edge-orientation entropy, variances of neural network features, and color statistics) for each image. In a rating study, we asked participants to evaluate the images along three aesthetic dimensions (Pleasing, Harmonious, and Interesting). Results demonstrate that not only objective image properties, but also subjective aesthetic preferences transferred from the original artworks onto the styletransferred images. The image properties of the style-transferred images explain 50-69% of the variance in the ratings. In the multidimensional space of statistical image properties, participants considered style-transferred images to be more Pleasing and Interesting if they were closer to a "sweet spot" where traditional Western paintings (JenAesthetics dataset) are represented. We conclude that NST is a useful tool to create novel artistic stimuli that preserve the image properties of the input style images. In the novel stimuli, we found a strong relationship between statistical image properties and subjective ratings, suggesting a prominent role of perceptual processing in the aesthetic evaluation of abstract images.
A model of aesthetic appreciation and aesthetic judgments
British Journal of Psychology, 2004
Although aesthetic experiences are frequent in modern life, there is as of yet no scientifically comprehensive theory that explains what psychologically constitutes such experiences. These experiences are particularly interesting because of their hedonic properties and the possibility to provide self-rewarding cognitive operations. We shall explain why modern art's large number of individualized styles, innovativeness and conceptuality offer positive aesthetic experiences. Moreover, the challenge of art is mainly driven by a need for understanding. Cognitive challenges of both abstract art and other conceptual, complex and multidimensional stimuli require an extension of previous approaches to empirical aesthetics. We present an information-processing stage model of aesthetic processing. According to the model, aesthetic experiences involve five stages: perception, explicit classification, implicit classification, cognitive mastering and evaluation. The model differentiates between aesthetic emotion and aesthetic judgments as two types of output.
Evaluation of paintings: Effects of lectures
Psihologija, 2014
This study investigated the influence of lectures about the Renaissance and abstract art on ratings of paintings from these two periods in art history. The study included two sessions. In the first, 72 naive participants rated the representational and abstract paintings. In the second session participants were divided into three groups: one received a lecture on Renaissance art, one attended a lecture on abstract art, and one group attended no lecture. Afterwards, the three groups rated a new, parallel set of paintings. Three first-order factors were extracted: Aesthetic experience, Relaxation tone, and Arousal. However, the higherorder General Aesthetic Experience factor explained a much higher amount of variance than the first-order factors, indicating its strong and generalized influence on naïve participants' experience with artworks. After the lecture on abstract art the participants rated paintings, especially abstract, as more aesthetically pleasing than the participants who attended the lecture on Renaissance art or the group without a lecture. Proposed explanation for this is that the naïve observers` ratings of abstract paintings are more susceptible to the influence of style-related information. When rating abstract artwork naïve observers may be significantly influenced by additional information gathered outside of the artwork.
PLOS ONE, 2015
Across cultures and throughout recorded history, humans have produced visual art. This raises the question of why people report such an emotional response to artworks and find some works more beautiful or compelling than others. In the current study we investigated the interplay between art expertise, and emotional and preference judgments. Sixty participants (40 novices, 20 art experts) rated a set of 150 abstract artworks and portraits during two occasions: in a laboratory setting and in a museum. Before commencing their second session, half of the art novices received a brief training on stylistic and art historical aspects of abstract art and portraiture. Results showed that art experts rated the artworks higher than novices on aesthetic facets (beauty and wanting), but no group differences were observed on affective evaluations (valence and arousal). The training session made a small effect on ratings of preference compared to the non-trained group of novices. Overall, these findings are consistent with the idea that affective components of art appreciation are less driven by expertise and largely consistent across observers, while more cognitive aspects of aesthetic viewing depend on viewer characteristics such as art expertise.
Individual differences as predictors of preference in visual art
Journal of Personality, 1985
A study was conducted to determine the value of personality and background vanables (e g , emotional responsivity, anxiety, parental memones, art training) as predictors of painting preferences One set of 24 paintings was selected, holding theme (sexual or aggressive) constant while contrasting Idealized (12) with Expressionist (12) styles Another set contrasted paintings on the Representational ys Abstract and Linear vs Painterly stylistic dimensions Contrasting pairs of paintings were presented to 48 subjects who rated preference under two instructional sets (objective interest and subjective pleasing) Questionnaire measures of personality and background were entered into step-wise multiple regression analyses, one for each type of preference and each stimulus contrast (8 altogether) Affective characteristics of the viewer (e g , anxiety, emotional responsivity) and parental memories (e g , of maternal affection or expressivity) predicted subjective preference For example, subjects who remembered their mothers as either expressive or affectionate preferred Idealized versions of sexual and aggressive paintings Artistic training and background emerged only in relation to objective preferences
Computational Intelligence and Neuroscience, 2018
How to interpret the relationship between the low-level features, such as some statistical characteristics of color and texture, and the high-level aesthetic properties, such as warm or cold, soft or hard, has been a hot research topic of neuroaesthetics. Contrary to the black-box method widely used in the fields of machine learning and pattern recognition, we build a white-box model with the hierarchical feed-forward structure inspired by neurobiological mechanisms underlying the aesthetic perception of visual art. In the experiment, the aesthetic judgments for 8 pairs of aesthetic antonyms are carried out for a set of 151 visual textures. For each visual texture, 106 low-level features are extracted. Then, ten more useful and effective features are selected through neighborhood component analysis to reduce information redundancy and control the complexity of the model. Finally, model building of the beauty appreciation of visual textures using multiple linear or nonlinear regressi...