Deep high dynamic range imaging of dynamic scenes

By Nima Khademi Kalantari, Ravi Ramamoorthi. ACM Transactions on Graphics, 2017

617 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/3c012139b2f01eae56931ab7cfde51dfee08f37c

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

End-to-end optimization of optics and image processing for achromatic extended depth of field and super-resolution imaging (2018, background): “For example, high dynamic range image estimation from a single low dynamic range photograph was recently demonstrated to achieve unprecedented image quality [Eilertsen et al. 2017; Kalantari and Ramamoorthi 2017; Zhang and Lalonde 2017].”
Deep High Dynamic Range Imaging with Large Foreground Motions (2017, methodology): “In [14], they first used optical flow to align input LDR images, followed by feeding the aligned LDRs into a convolutional neural network (CNN) to produce the final HDR image.”
ExpandNet: A Deep Convolutional Neural Network for High Dynamic Range Expansion from Low Dynamic Range Content (2018, background): “…predicting sky illumination for rendering [SBRCD17, HGSH ∗ 17], denoising Monte Carlo renderings [KBS15, CKS ∗ 17, BVM ∗ 17], predicting HDR environment maps [ZL17a], reducing artefacts such as ghosting when fusing multiple LDR exposures to create HDR content [KR17] and for tone mapping [HDQ17].”
Learning a Deep Single Image Contrast Enhancer from Multi-Exposure Images (2018, background): “In the last decade, researchers have spent much efforts to design de-ghosting algorithms [2], [14], [32] and learning-based methods like [33] proposed to map the multi-exposure image sequences to an HDR image.”
Generative Diffusion Prior for Unified Image Restoration and Enhancement (2023, methodology): “The stateof-the-art methods used for comparison include AHDRNet [86], HDR-GAN [56], DeepHDR [85] and deep-highdynamic-range [29].”
Attention-Guided Network for Ghost-Free High Dynamic Range Imaging (2019, methodology): “Previous methods [15, 37] take stacked LDR images, or LDR image feature maps, as the input to the merging process, which mixes the misaligned image components at an early stage of the network, making it difficult to obtain ghost-free HDR results.”
Deep blending for free-viewpoint image-based rendering (2018, methodology): “…reined geometry to provide ranked source-image mosaics to the CNN. Finally, deep learning has been used to estimate weights for HDR blending [Kalantari and Ramamoorthi 2017], or for iltering for Monte-Carlo path tracing [Bako et al. 2017]; while these methods are similar in spirit to our…”
HDR image reconstruction from a single exposure using deep CNNs (2017, methodology): “Recently, deep learning has also been successfully applied for improving classical HDR video reconstruction from multiple exposures captured over time [Kalantari and Ramamoorthi 2017].”
EnlightenGAN: Deep Light Enhancement Without Paired Supervision (2019, methodology): “It also avoids overﬁtting any speciﬁc data generation protocol or imaging device that previous works [15, 5, 16] implicitly rely on, hence leading to notably improved real-world generalization.”