Pinned Repositories
ArtCoder
A PyTorch implementation of "ArtCoder: An End-to-end Method for Generating Scanning-robust Stylized QR Codes", CVPR 2021.
LandmarkTool
A tool for labeling 68 face landmarks based on python-Qt5.
Mang2Vec
MARVEL: Raster Gray-level Manga Vectorization via Primitive-wise Deep Reinforcement Learning
neural-style-pytorch
A fast PyTorch implementation of "A Neural Algorithm of Artistic Style"
Videolized-Aesthetic-QR-Codes
Quick Response (QR) code is one of the most worldwide used two-dimensional codes. Traditional QR codes consist of visual-unpleasant black/white square modules, which triggers various existing works of image-pattern Aesthetic QR Codes (AQRCs) to beautify the appearances of such codes. In this paper, we innovatively propose the Videolized-AQRCs (V-AQRCs) that can blend videos with QR codes adaptively while preserving the machine-readability and the video content. In our approach, firstly, we make QR codes' modules adaptive to video content to enhance the visual quality by utilizing the contextual information of the video. Secondly, an effective algorithm is designed to optimize the serious flickering of adaptive modules. Finally, we naturally blend modules with frames by balancing the visual effect and readability, and then synthesize the robust V-AQRCs. Experimental results show that our V-AQRCs have practical value, and high qualities in terms of visual attraction and robustness, which can support the real-world application.
Videolized_Aesthetic_QR_codes
SwordHolderSH's Repositories
SwordHolderSH/ArtCoder
A PyTorch implementation of "ArtCoder: An End-to-end Method for Generating Scanning-robust Stylized QR Codes", CVPR 2021.
SwordHolderSH/Mang2Vec
MARVEL: Raster Gray-level Manga Vectorization via Primitive-wise Deep Reinforcement Learning
SwordHolderSH/neural-style-pytorch
A fast PyTorch implementation of "A Neural Algorithm of Artistic Style"
SwordHolderSH/LandmarkTool
A tool for labeling 68 face landmarks based on python-Qt5.
SwordHolderSH/Videolized-Aesthetic-QR-Codes
Quick Response (QR) code is one of the most worldwide used two-dimensional codes. Traditional QR codes consist of visual-unpleasant black/white square modules, which triggers various existing works of image-pattern Aesthetic QR Codes (AQRCs) to beautify the appearances of such codes. In this paper, we innovatively propose the Videolized-AQRCs (V-AQRCs) that can blend videos with QR codes adaptively while preserving the machine-readability and the video content. In our approach, firstly, we make QR codes' modules adaptive to video content to enhance the visual quality by utilizing the contextual information of the video. Secondly, an effective algorithm is designed to optimize the serious flickering of adaptive modules. Finally, we naturally blend modules with frames by balancing the visual effect and readability, and then synthesize the robust V-AQRCs. Experimental results show that our V-AQRCs have practical value, and high qualities in terms of visual attraction and robustness, which can support the real-world application.
SwordHolderSH/Videolized_Aesthetic_QR_codes