Zobrazeno 1 - 10
of 24
pro vyhledávání: '"Wang, Zhouxia"'
Although learning-based image restoration methods have made significant progress, they still struggle with limited generalization to real-world scenarios due to the substantial domain gap caused by training on synthetic data. Existing methods address
Externí odkaz:
http://arxiv.org/abs/2406.18516
Diffusion model has demonstrated remarkable capability in video generation, which further sparks interest in introducing trajectory control into the generation process. While existing works mainly focus on training-based methods (e.g., conditional ad
Externí odkaz:
http://arxiv.org/abs/2406.16863
Autor:
Li, Yaowei, Wang, Xintao, Zhang, Zhaoyang, Wang, Zhouxia, Yuan, Ziyang, Xie, Liangbin, Zou, Yuexian, Shan, Ying
Filmmaking and animation production often require sophisticated techniques for coordinating camera transitions and object movements, typically involving labor-intensive real-world capturing. Despite advancements in generative AI for video creation, a
Externí odkaz:
http://arxiv.org/abs/2406.15339
Recovering degraded low-resolution text images is challenging, especially for Chinese text images with complex strokes and severe degradation in real-world scenarios. Ensuring both text fidelity and style realness is crucial for high-quality text ima
Externí odkaz:
http://arxiv.org/abs/2312.08886
Autor:
Wang, Zhouxia, Yuan, Ziyang, Wang, Xintao, Chen, Tianshui, Xia, Menghan, Luo, Ping, Shan, Ying
Motions in a video primarily consist of camera motion, induced by camera movement, and object motion, resulting from object movement. Accurate control of both camera and object motion is essential for video generation. However, existing works either
Externí odkaz:
http://arxiv.org/abs/2312.03641
Autor:
Wang, Zhouxia, Wang, Xintao, Xie, Liangbin, Qi, Zhongang, Shan, Ying, Wang, Wenping, Luo, Ping
This paper presents a LoRA-free method for stylized image generation that takes a text prompt and style reference images as inputs and produces an output image in a single pass. Unlike existing methods that rely on training a separate LoRA for each s
Externí odkaz:
http://arxiv.org/abs/2309.01770
Blind face restoration aims at recovering high-quality face images from those with unknown degradations. Current algorithms mainly introduce priors to complement high-quality details and achieve impressive progress. However, most of these algorithms
Externí odkaz:
http://arxiv.org/abs/2308.07228
Blind face restoration is to recover a high-quality face image from unknown degradations. As face image contains abundant contextual information, we propose a method, RestoreFormer, which explores fully-spatial attentions to model contextual informat
Externí odkaz:
http://arxiv.org/abs/2201.06374
Automatically selecting exposure bracketing (images exposed differently) is important to obtain a high dynamic range image by using multi-exposure fusion. Unlike previous methods that have many restrictions such as requiring camera response function,
Externí odkaz:
http://arxiv.org/abs/2005.12536