Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features
In this paper, we explore how higher-level perceptual information based on visual attention can be used for aesthetic assessment of images. We assume that visually dominant subjects in a photograph influence stronger aesthetic interest. In other words, visual attention may be a key to predicting pho...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2021-01-01
|
| Series: | Applied Artificial Intelligence |
| Online Access: | http://dx.doi.org/10.1080/08839514.2020.1839197 |
| _version_ | 1797684889282150400 |
|---|---|
| author | Hironori Takimoto Fumiya Omori Akihiro Kanagawa |
| author_facet | Hironori Takimoto Fumiya Omori Akihiro Kanagawa |
| author_sort | Hironori Takimoto |
| collection | DOAJ |
| description | In this paper, we explore how higher-level perceptual information based on visual attention can be used for aesthetic assessment of images. We assume that visually dominant subjects in a photograph influence stronger aesthetic interest. In other words, visual attention may be a key to predicting photographic aesthetics. Our proposed aesthetic assessment method, which is based on multi-stream and multi-task convolutional neural networks (CNNs), extracts global features and saliency features from an input image. These provide higher-level visual information such as the quality of the photo subject and the subject–background relationship. Results from our experiments support the effectiveness of our approach. |
| first_indexed | 2024-03-12T00:36:15Z |
| format | Article |
| id | doaj.art-8a15bf178d0f4c4ea3245a288151d8f7 |
| institution | Directory Open Access Journal |
| issn | 0883-9514 1087-6545 |
| language | English |
| last_indexed | 2024-03-12T00:36:15Z |
| publishDate | 2021-01-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Applied Artificial Intelligence |
| spelling | doaj.art-8a15bf178d0f4c4ea3245a288151d8f72023-09-15T09:33:58ZengTaylor & Francis GroupApplied Artificial Intelligence0883-95141087-65452021-01-01351254010.1080/08839514.2020.18391971839197Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency FeaturesHironori Takimoto0Fumiya Omori1Akihiro Kanagawa2Okayama Prefectural UniversityOkayama Prefectural UniversityOkayama Prefectural UniversityIn this paper, we explore how higher-level perceptual information based on visual attention can be used for aesthetic assessment of images. We assume that visually dominant subjects in a photograph influence stronger aesthetic interest. In other words, visual attention may be a key to predicting photographic aesthetics. Our proposed aesthetic assessment method, which is based on multi-stream and multi-task convolutional neural networks (CNNs), extracts global features and saliency features from an input image. These provide higher-level visual information such as the quality of the photo subject and the subject–background relationship. Results from our experiments support the effectiveness of our approach.http://dx.doi.org/10.1080/08839514.2020.1839197 |
| spellingShingle | Hironori Takimoto Fumiya Omori Akihiro Kanagawa Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features Applied Artificial Intelligence |
| title | Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features |
| title_full | Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features |
| title_fullStr | Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features |
| title_full_unstemmed | Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features |
| title_short | Image Aesthetics Assessment Based on Multi-stream CNN Architecture and Saliency Features |
| title_sort | image aesthetics assessment based on multi stream cnn architecture and saliency features |
| url | http://dx.doi.org/10.1080/08839514.2020.1839197 |
| work_keys_str_mv | AT hironoritakimoto imageaestheticsassessmentbasedonmultistreamcnnarchitectureandsaliencyfeatures AT fumiyaomori imageaestheticsassessmentbasedonmultistreamcnnarchitectureandsaliencyfeatures AT akihirokanagawa imageaestheticsassessmentbasedonmultistreamcnnarchitectureandsaliencyfeatures |