958 Hits in 5.5 sec

Ground Truth Free Denoising by Optimal Transport [article]

Sören Dittmer, Carola-Bibiane Schönlieb, Peter Maass
2020 arXiv   pre-print
We present a learned unsupervised denoising method for arbitrary types of data, which we explore on images and one-dimensional signals.  ...  Ground truth free denoising via dual critics We will now formalize the main idea of the paper.  ...  CONCLUSION This paper introduces a novel way of training a denoiser without any kind of ground truth data by making use of optimal transport, more specifically by using a modified Wasserstein GAN setting  ... 
arXiv:2007.01575v1 fatcat:k2r4rdewyndrdgyvudbiwtaqvu

DeepMCDose: A Deep Learning Method for Efficient Monte Carlo Beamlet Dose Calculation by Predictive Denoising in MR-Guided Radiotherapy [article]

Ryan Neph, Yangsibo Huang, Youming Yang, Ke Sheng
2019 arXiv   pre-print
Our model achieves a normalized mean absolute error of only 0.106% compared with the ground truth dose contrasting the 25.7% error of the under sampled MC dose fed into the network at prediction time.  ...  Our method uses parallel UNET branches acting on three input channels before mixing latent understanding to produce noise-free dose predictions.  ...  Miao [7] investigated the use of an adaptive denoising approach modeling the dose in terms of heat transport and used anisotropic diffusion to achieve smoothed distributions.  ... 
arXiv:1908.04437v1 fatcat:guvjcbkaojbdjmsoj47lnezjxq

A survey on deep learning-based Monte Carlo denoising

Yuchi Huo, Sung-eui Yoon
2021 Computational Visual Media  
Recent years have seen increasing attention and significant progress in denoising MC rendering with deep learning, by training neural networks to reconstruct denoised rendering results from sparse MC samples  ...  MC denoising can be formally described as a mapping g of an input x to the ground-truth r rendered by a high sample count (Fig. 1 ).  ...  between the reconstructed output and the ground truth.  ... 
doi:10.1007/s41095-021-0209-9 fatcat:pki3mbbpw5fjxf6oaphcnlto4a

N-SfC: Robust and Fast Shape Estimation from Caustic Images [article]

Marc Kassubeck, Moritz Kappel, Susana Castillo, Marcus Magnor
2021 arXiv   pre-print
the computational cost of the light transport simulation, and an optimization process based on learned gradient descent, which enables better convergence using fewer iterations.  ...  The recent Shape from Caustics (SfC) method casts the problem as the inverse of a light propagation simulation for synthesis of the caustic image, that can be solved by a differentiable renderer.  ...  Denoising dataset: we generate training data for our denoiser by sampling random ground-truth height fields from line distribution (top), which are then used to render low quality (n l =10 6 samples, middle  ... 
arXiv:2112.06705v1 fatcat:vkrok2rkv5h4doczc4olpeai5i

A detail preserving neural network model for Monte Carlo denoising

Weiheng Lin, Beibei Wang, Lu Wang, Nicolas Holzschuch
2020 Computational Visual Media  
In this paper, we solve this issue by proposing a novel network structure, a new input feature-light transport covariance from path space-and an improved loss function.  ...  In addition, we use a light transport covariance feature in path space as one of the features, which helps to preserve illumination details.  ...  between the filtered color and ground truth.  ... 
doi:10.1007/s41095-020-0167-7 fatcat:cozemwjkkzbwdc2hkwlzyalwxe

DEMC: A Deep Dual-Encoder Network for Denoising Monte Carlo Rendering [article]

Xin Yang, Wenbo Hu, Dawei Wang, Lijing Zhao, Baocai Yin, Qiang Zhang, Xiaopeng Wei, Hongbo Fu
2019 arXiv   pre-print
Denoising Monte Carlo rendering is different from natural image denoising since inexpensive by-products (feature buffers) can be extracted in the rendering stage.  ...  Most of them are noise-free and can provide sufficient details for image reconstruction.  ...  [24] applied a linear model to approximate the ground truth, by weighting the error of each pixel based on the auxiliary features. Bitterli et al.  ... 
arXiv:1905.03908v1 fatcat:65sdtdfmxvdjvkmcroysjh7ali

Highway Elevation Data Smoothing Using Local Enhancement Mechanism and Butterworth Filter

Xinqiang Chen, Zhibin Li, Yinhai Wang, Chaojian Shi, Huafeng Wu, Shengzheng Wang
2017 International Journal of Innovative Computing, Information and Control  
Google Earth provides public accessible highway elevation data which is deemed as a credible elevation source by many researchers and transportation practitioners.  ...  Segment's elevation under local extrema grades was adjusted by the local enhancement mechanism before the Butterworth smoothing process.  ...  As freeway street view from Google Map was recorded by onboard cameras, the images of buckling surface were taken as ground truth images.  ... 
doi:10.24507/ijicic.13.06.1887 fatcat:bsqnqvph2jdtrdgzhl4nsskwae

Deep residual learning for denoising Monte Carlo renderings

Kin-Ming Wong, Tien-Tsin Wong
2019 Computational Visual Media  
We have evaluated our method on unseen images created by a different renderer. Consistently superior quality denoising is obtained in all cases.  ...  In this paper, we propose a deep residual learning based method that outperforms both state-of-the-art handcrafted denoisers and learning-based denoisers.  ...  Figure 6 shows selected ground truth images from our dataset.  ... 
doi:10.1007/s41095-019-0142-3 fatcat:wcq3vsk7cnce3ome464bh5izse

Physically Consistent and Efficient Variational Denoising of Image Fluid Flow Estimates

A. Vlasenko, C. Schnorr
2010 IEEE Transactions on Image Processing  
The results illustrates that about 30% of TV-denoised vectors have directions opposite to the ground truth flow, while the result returned by our method nearly matches ground truth.  ...  Middle: Ground truth. Bottom: Vector field denoised with = 0:1.  ...  His research interests include image processing, computer vision and pattern analysis, and corresponding problems of mathematical modeling and optimization.  ... 
doi:10.1109/tip.2009.2036673 pmid:19933005 fatcat:daooazj5jfalbdvuxzfehh4fae

Learning Neural Light Transport [article]

Paul Sanzenbacher, Lars Mescheder, Andreas Geiger
2020 arXiv   pre-print
Moreover, it compares favorably to baselines which combine path tracing and image denoising at the same computational budget.  ...  We present an approach for learning light transport in static and dynamic 3D scenes using a neural network with the goal of predicting photorealistic images.  ...  Predictions and error images with respect to ground truth for different denoising approaches and our approach for dynamic objects and fixed lights.  ... 
arXiv:2006.03427v1 fatcat:4knpevlvljdtjih4ne3liynlfy

Real-Time Denoising of Volumetric Path Tracing for Direct Volume Rendering

Jose A. Iglesias-Guitian, Prajita Sukhdev Mane, Bochang Moon
2020 IEEE Transactions on Visualization and Computer Graphics  
This makes existing real-time denoisers, which take noise-free G-buffers as their input, less effective when denoising VPT images.  ...  In particular, our denoising exploits temporal coherence between frames, without relying on noise-free G-buffers, which has been a common assumption of existing denoisers for surface-models.  ...  Ideally, for a given pixel j at frame t, we would compute the real error by using the values of the ground truth image I j , as e j (t) = I j (t)−Î j (t).  ... 
doi:10.1109/tvcg.2020.3037680 pmid:33180727 fatcat:vrlecl52jncqhmazy46ntdagkq

Real-time Monte Carlo Denoising with the Neural Bilateral Grid

Xiaoxu Meng, Quan Zheng, Amitabh Varshney, Gurprit Singh, Matthias Zwicker
2020 Eurographics Symposium on Rendering  
This causes severe noise, and previous techniques strongly compromise final quality to maintain real-time denoising speed.  ...  In addition, we also show how we can further improve denoising quality using a hierarchy of multi-scale bilateral grids.  ...  ., ((r N , f N ),t N )}, where r i , f i stands for noisy radiance and auxiliary features of frame i, and t i is the noise-free ground truth of frame i.  ... 
doi:10.2312/sr.20201133 dblp:conf/rt/MengZVSZ20 fatcat:qpzyz3eihfgmfmp6ko7zgbtmhy

Listening to Sounds of Silence for Speech Denoising [article]

Ruilin Xu, Rundi Wu, Yuko Ishiwaka, Carl Vondrick, Changxi Zheng
2020 arXiv   pre-print
Experiments on multiple datasets confirm the pivotal role of silent interval detection for speech denoising, and our method outperforms several state-of-the-art denoising methods, including those that  ...  We introduce a deep learning model for speech denoising, a long-standing challenge in audio analysis arising in numerous applications.  ...  This work was supported in part by the National Science Foundation (1717178, 1816041, 1910839, 1925157) and SoftBank Group.  ... 
arXiv:2010.12013v1 fatcat:zexcxcyf5nhs7kyhd6ltxz23ve

Survey: Machine Learning in Production Rendering [article]

Shilin Zhu
2020 arXiv   pre-print
light transport situations.  ...  Some of these techniques have already been used in the latest released animations while others are still in the continuing development by researchers in both academia and movie studios.  ...  However, rendering high-sample-count ground-truth reference images was time-consuming and prevented the dataset from scaling up.  ... 
arXiv:2005.12518v1 fatcat:htymxgpc2jdmre6umrlh2xvsu4

Path Tracing Denoising based on SURE Adaptive Sampling and Neural Network

Qiwei Xing, Chunyi Chen
2020 IEEE Access  
An anisotropic filter is used to reconstruct the final images with the parameters predicted by neural networks.  ...  INDEX TERMS Adaptive sampling, SURE estimator, MLPs network, path tracing, denoising.  ...  Note the ground truth image still has visible spikes at 96K spp, while we produce a relatively noise-free result.  ... 
doi:10.1109/access.2020.2999891 fatcat:l3efvetnefbptez3ejknvucvgy
« Previous Showing results 1 — 15 out of 958 results