ILGNet: inception modules with connected local and global features for efficient image aesthetic quality classification using domain adaptation.
Xin JinLe WuXiaodong LiXiaokun ZhangJingying ChiSiwei PengShiming GeGeng ZhaoShuying LiPublished in: IET Comput. Vis. (2019)
Keyphrases
- global features
- domain adaptation
- feature vectors
- image features
- image classification
- classification method
- visual features
- input image
- feature representation
- multiscale
- keypoints
- image content
- image retrieval
- image representation
- classification accuracy
- feature space
- support vector machine
- information retrieval
- cross domain
- classification algorithm
- decision trees
- document classification
- multiple sources
- feature extraction
- support vector
- text classification
- test data
- labeled data
- k nearest neighbor
- low level
- image processing