2017 여름방학

Neural style (CNN training)

• http://www.cv-foundation.org/openaccess/content_cvpr_2016/papers/Gatys_Image_Style_Transfer_CVPR_2016_paper.pdf
• The first image style transferring (image to image)

Deep Photo style transfer

• https://arxiv.org/abs/1703.07511
• upgrade Neural style by photo realism regularization

Variational Auto Encoder (VAE)

• https://arxiv.org/abs/1312.6114
• The first feedforward network for image style transferring (style learning and fast generation)

Importance Weighted Auto Encoder

• https://arxiv.org/abs/1509.00519
• upgrade VAE by importance weighting

Perceptual Loss + Super resolution

• https://arxiv.org/abs/1603.08155
• combine Neural style (conceptual loss) and fast feedforward

Multi-style in a network

• https://arxiv.org/abs/1610.07629
• upgrade Perceptual Loss + Super resolution method by adding multi-style network instead of single style network

Conceptual factors of image style

• https://arxiv.org/abs/1611.07865
• conceptual factor analysis from Gram style feature

Arbitrary Style Transfer

• https://arxiv.org/abs/1703.06868v2
• adaptive instance normalization

image to image translation

• https://arxiv.org/abs/1611.07004
• conditional gan

real time object detector(YOLO)

• https://arxiv.org/abs/1506.02640
• unified object detector

Notice (2019.8.9)

• To be updated with tensorflow r1.14