Research Article
Style Transfer of Chinese Art Works Based on Dual Channel Deep Learning Model
Table 1
Parameters of art works data set prepared.
| Artwork dataset name | Positive sample (%) | Size | Resolution (dpi) |
| Labeled faces in the wild | 48273 (7.64) | 640 × 480 | 100 million | MNIST-06 | 39390 (39.8) | 1280 × 640 | 20 million | CIFAR-02 | 36640 (8.21) | 640 × 480 | 100 million | AI-challneger-04 | 63605 (40.2) | 1280 × 640 | 20 million | Pascal VOC-07 | 27001 (47.21) | 640 × 480 | 100 million | COCO common objects dataset | 75220 (4.1) | 1280 × 640 | 20 million | CityScapes | 51220 (46.9) | 640 × 480 | 100 million | Lego bricks | 19090 (38.67) | 640 × 480 | 100 million | Visual genome | 75904 (17.62) | 640 × 480 | 100 million | VisualQA | 72490 (25.05) | 1280 × 640 | 20 million | MNIST-08 | 36257 (47.14) | 640 × 480 | 100 million | KITTI-05 | 28284 (30.13) | 1280 × 640 | 20 million | ApolloScape-02 | 33298 (18.37) | 1280 × 640 | 20 million | TUM-08 | 68273 (1.9) | 640 × 480 | 100 million |
|
|