Shusheng Puxiang - A high dynamic range imaging algorithm jointly developed by Shanghai AI Lab, the Chinese University of Hong Kong, and Zhejiang University.
The "Shusheng·Puxiang" algorithm is an ultra-high dynamic range imaging algorithm developed by the Shanghai Artificial Intelligence Laboratory in collaboration with the Chinese University of Hong Kong, Zhejiang University, and other institutions. Its core is Puxiang HDR (UltraFusion HDR). The algorithm combines AIGC technology and HDR technology...
What is the scholar Urakashi?
Shusheng·Puxiang is an ultra-high dynamic range imaging algorithm developed by the Shanghai Artificial Intelligence Laboratory in collaboration with the Chinese University of Hong Kong, Zhejiang University, and other institutions. Its core is Puxiang HDR (UltraFusion HDR). The algorithm combines AIGC and HDR technologies, utilizing exposure fusion and a generative large model to restore image details and generate high-quality images even under extreme conditions with exposure differences of up to 9 stops. It effectively solves the motion artifact problem in dynamic scenes and is applicable to fields such as photography, mobile imaging, and autonomous driving.
The main functions of the scholar-portrait
- Ultra-high dynamic range imagingBy combining AIGC technology with HDR, it is possible to restore image details and generate high-quality images under extreme conditions with exposure differences of up to 9 stops.
- Dynamic scene processingIt effectively solves the problem of motion artifacts in dynamic objects and is suitable for imaging under complex lighting conditions.
- Multi-exposure fusionIt supports arbitrary exposure input and can merge images with different exposures into a single high-quality image with vibrant colors and rich details.
The technical principles of the scholar Urakashi
- Multi-exposure fusion and generative large modelsBased on AIGC technology, Uraraka HDR can process images with arbitrary exposure input. Even under extreme conditions with exposure differences of up to 9 stops, it can still restore image details and generate high-quality HDR images.
- Dynamic scene processingTo address motion artifacts in dynamic scenes, Puxiang HDR employs an innovative multi-stage processing workflow to effectively solve imaging problems caused by the movement of dynamic objects, ensuring that the generated images are natural and realistic.
- Tone mapping and image restorationThe algorithm overcomes the unnatural tones of traditional HDR techniques under complex lighting conditions by adaptively learning tone mapping through prior knowledge from generative large models. The guided image inpainting design framework maintains the fidelity of the generated images and avoids texture distortion issues.
- AIGC technology empowermentAIGC technology provides powerful generation capabilities for HDR imaging, enabling algorithms to achieve efficient fusion between different exposure levels, significantly improving the robustness and color saturation of the image.
The project address of Shusheng·Puxiang
- Official Homepage:https://ultrafusion.openxlab.org.cn/home
- Project official website:https://openimaginglab.github.io/UltraFusion/
- Experience the demo online:https://huggingface.co/spaces/iimmortall/UltraFusion
Application scenarios of Scholar Pu Xiang
- PhotographyIt helps photographers repair underexposed or overexposed photos, improving image quality without upgrading equipment.
- Mobile phone photographyIt is expected to be integrated into mobile phone cameras to improve the imaging performance of mobile phones under complex lighting conditions.
- autonomous drivingIn autonomous driving systems, HDR from Urahashi can better process images in complex environments, improving the system's perception capabilities and safety.