Pleasant/Unpleasant Filtering for Affective Image Retrieval Based on Cross-Correlation of EEG Features

Abstract

People often make decisions based on sensitivity rather than rationality. In the field of biological information processing, methods are available for analyzing biological information directly based on electroencephalogram: EEG to determine the pleasant/unpleasant reactions of users. In this study, we propose a sensitivity filtering technique for discriminating preferences (pleasant/unpleasant) for images using a sensitivity image filtering system based on EEG. Using a set of images retrieved by similarity retrieval, we perform the sensitivity-based pleasant/unpleasant classification of images based on the affective features extracted from images with the maximum entropy method: MEM. In the present study, the affective features comprised cross-correlation features obtained from EEGs produced when an individual observed an image. However, it is difficult to measure the EEG when a subject visualizes an unknown image. Thus, we propose a solution where a linear regression method based on canonical correlation is used to estimate the cross-correlation features from image features. Experiments were conducted to evaluate the validity of sensitivity filtering compared with image similarity retrieval methods based on image features. We found that sensitivity filtering using color correlograms was suitable for the classification of preferred images, while sensitivity filtering using local binary patterns was suitable for the classification of unpleasant images. Moreover, sensitivity filtering using local binary patterns for unpleasant images had a 90% success rate. Thus, we conclude that the proposed method is efficient for filtering unpleasant images.

Electroencephalography Amplitude Modulation Analysis for Automated Affective Tagging of Music Video Clips

“…While the AMI features capture non-linear interactions between inter-electrode amplitude-amplitude coupling patterns, the Pearson correlation coefficient between the patterns can also be used to quantify the coherence, or linear interactions between the patterns. Spectral coherence measures have been widely used in EEG research and were recently shown to also be useful for affective state research (e.g., Kar et al, 2014 ; Xielifuguli et al, 2014 ). Hence, we explore the concept of amplitude modulation coherence, or AMC as a new feature for affective state recognition.…”

On the Influence of Affect in EEG-Based Subject Identification

“…This translates into the necessity that samples that belong to the same individual appear close to each other, according to a concrete distance and in relation to samples coming from different subjects. To test if this condition holds for our data, we have computed the cosine distance (D C ) between all pairs of samples in each dataset, in line with other research works that have remarked on the benefits of this distance in specific contexts [50], [51]. This has been computed as:…”