Abstract | ||
---|---|---|
This paper addresses the problem of Visual Sentiment Analysis focusing on the estimation of the polarity of the sentiment evoked by an image. Starting from an embedding approach which exploits both visual and textual features, we attempt to boost the contribution of each input view. We propose to extract and employ anObjective Textdescription of images rather than the classicSubjective Textprovided by the users (i.e., title, tags and image description) which is extensively exploited in the state of the art to infer the sentiment associated to social images.Objective Textis obtained from the visual content of the images through recent deep learning architectures which are used to classify object, scene and to perform image captioning.Objective Textfeatures are then combined with visual features in an embedding space obtained with Canonical Correlation Analysis. The sentiment polarity is then inferred by a supervised Support Vector Machine. During the evaluation, we compared an extensive number of text and visual features combinations and baselines obtained by considering the state of the art methods. Experiments performed on a representative dataset of 47235 labelled samples demonstrate that the exploitation ofObjective Texthelps to outperform state-of-the-art for sentiment polarity estimation. |
Year | DOI | Venue |
---|---|---|
2021 | 10.1007/s11042-019-08312-7 | MULTIMEDIA TOOLS AND APPLICATIONS |
Keywords | DocType | Volume |
Visual sentiment analysis, Objective text features, Embedding spaces, Social media | Journal | 80 |
Issue | ISSN | Citations |
15 | 1380-7501 | 0 |
PageRank | References | Authors |
0.34 | 0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
alessandro ortis | 1 | 24 | 8.54 |
Giovanni Maria Farinella | 2 | 412 | 57.13 |
Giovanni Torrisi | 3 | 15 | 1.79 |
Sebastiano Battiato | 4 | 20 | 6.74 |