High Dynamic Range

High dynamic range (HDR) imaging aims to capture and reproduce the full range of light intensities present in a scene, exceeding the limitations of standard cameras. Current research focuses on developing efficient algorithms and neural network architectures, such as diffusion models and Gaussian splatting, to reconstruct HDR images and videos from multiple low dynamic range (LDR) exposures or single LDR images, often incorporating techniques like tone mapping and exposure completion to address challenges such as ghosting and limited dynamic range. These advancements are significant for improving image and video quality in various applications, including computer vision, virtual and augmented reality, and advanced driver-assistance systems.