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 namePositive sample (%)SizeResolution (dpi)

Labeled faces in the wild48273 (7.64)640 × 480100 million
MNIST-0639390 (39.8)1280 × 64020 million
CIFAR-0236640 (8.21)640 × 480100 million
AI-challneger-0463605 (40.2)1280 × 64020 million
Pascal VOC-0727001 (47.21)640 × 480100 million
COCO common objects dataset75220 (4.1)1280 × 64020 million
CityScapes51220 (46.9)640 × 480100 million
Lego bricks19090 (38.67)640 × 480100 million
Visual genome75904 (17.62)640 × 480100 million
VisualQA72490 (25.05)1280 × 64020 million
MNIST-0836257 (47.14)640 × 480100 million
KITTI-0528284 (30.13)1280 × 64020 million
ApolloScape-0233298 (18.37)1280 × 64020 million
TUM-0868273 (1.9)640 × 480100 million