Perception of noise in global illumination based on inductive learning

Constantin, Joseph; Constantin, Ibtissam; Bigand, André; Hamad, Denis

doi:10.1109/ijcnn.2016.7727861

Cited by 3 publications

(5 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Next, these models were compared with the standard support vector machine (SVM) and the active semi-supervised SVM called

by considering the full luminance component of

color sub-images ( L ) as input and applying to L four different denoising algorithms: Linear filtering with averaging and Gaussian filters, median filters and adaptive Wiener filters. Each sub-image was also denoised using Wavelet analysis [ 33 ]. A comparison has also been made with the pre-trained VGG19 deep convolution neural network in order to show the efficiency of our work.…”

Section: The Proposed Methodsmentioning

confidence: 99%

“…The semi-supervised learning algorithm is applied for the unlabeled remaining sub-images that are not chosen during active learning [ 16 ]. This algorithm adds class central sub-images to describe the class distribution into the labeled training set, then the class boundary could be very beneficial for accelerating the convergence [ 33 ].…”

Section: The Proposed Methodsmentioning

confidence: 99%

“…Other research papers combined active with semi-supervised learning for image compression, speech recognition, and multi-task learning [ 30 , 31 , 32 ]. A very competitive class of semi-supervised learning algorithms was proposed to exploit robustness in case of stochastic noise [ 33 ]. However, the proposed models take a full dimension of the image as input without using a deep convolution neural network (CNN), which is shown to be successful.…”

Section: Related Work On Active Semi-supervised Learningmentioning

confidence: 99%

See 2 more Smart Citations

On the Use of Deep Active Semi-Supervised Learning for Fast Rendering in Global Illumination

Constantin

Bigand

2020

J. Imaging

Self Cite

View full text Add to dashboard Cite

Convolution neural networks usually require large labeled data-sets to construct accurate models. However, in many real-world scenarios, such as global illumination, labeling data are a time-consuming and costly human intelligent task. Semi-supervised learning methods leverage this issue by making use of a small labeled data-set and a larger set of unlabeled data. In this paper, our contributions focus on the development of a robust algorithm that combines active and deep semi-supervised convolution neural network to reduce labeling workload and to accelerate convergence in case of real-time global illumination. While the theoretical concepts of photo-realistic rendering are well understood, the increased need for the delivery of highly dynamic interactive content in vast virtual environments has increased recently. Particularly, the quality measure of computer-generated images is of great importance. The experiments are conducted on global illumination scenes which contain diverse distortions. Compared with human psycho-visual thresholds, the good consistency between these thresholds and the learning models quality measures can been seen. A comparison has also been made with SVM and other state-of-the-art deep learning models. We do transfer learning by running the convolution base of these models over our image set. Then, we use the output features of the convolution base as input to retrain the parameters of the fully connected layer. The obtained results show that our proposed method provides promising efficiency in terms of precision, time complexity, and optimal architecture.

show abstract

“…Next, these models were compared with the standard support vector machine (SVM) and the active semi-supervised SVM called

by considering the full luminance component of

Section: The Proposed Methodsmentioning

confidence: 99%

Section: The Proposed Methodsmentioning

confidence: 99%

Section: Related Work On Active Semi-supervised Learningmentioning

confidence: 99%

See 1 more Smart Citation

On the Use of Deep Active Semi-Supervised Learning for Fast Rendering in Global Illumination

Constantin

Bigand

2020

J. Imaging

Self Cite

View full text Add to dashboard Cite

show abstract

“…These values form the 26 input noise attributes of a SVM model in order to classify image as noisy or not. Authors also proposed an approach [ 41 ] where the previously proposed filters are applied one after the other on the input image. The frequency bases are then extracted from this filtered image and a noise-free image is obtained.…”

Section: Previous Workmentioning

confidence: 99%

“…It is thus very different of well-known noise models and using available natural image databases is not possible. Then data that were used in [ 41 , 42 ] are still limited both in terms of image size and scene complexity.…”

Section: Subjective Dataset Collectionmentioning

confidence: 99%

Stopping Criterion during Rendering of Computer-Generated Images Based on SVD-Entropy

Buisine

Bigand

Synave

et al. 2021

Entropy

View full text Add to dashboard Cite

The estimation of image quality and noise perception still remains an important issue in various image processing applications. It has also become a hot topic in the field of photo-realistic computer graphics where noise is inherent in the calculation process. Unlike natural-scene images, however, a reference image is not available for computer-generated images. Thus, classic methods to assess noise quantity and stopping criterion during the rendering process are not usable. This is particularly important in the case of global illumination methods based on stochastic techniques: They provide photo-realistic images which are, however, corrupted by stochastic noise. This noise can be reduced by increasing the number of paths, as proved by Monte Carlo theory, but the problem of finding the right number of paths that are required in order to ensure that human observers cannot perceive any noise is still open. Until now, the features taking part in the human evaluation of image quality and the remaining perceived noise are not precisely known. Synthetic image generation tends to be very expensive and the produced datasets are high-dimensional datasets. In that case, finding a stopping criterion using a learning framework is a challenging task. In this paper, a new method for characterizing computational noise for computer generated images is presented. The noise is represented by the entropy of the singular value decomposition of each block composing an image. These Singular Value Decomposition (SVD)-entropy values are then used as input to a recurrent neural network architecture model in order to extract image noise and in predicting a visual convergence threshold of different parts of any image. Thus a new no-reference image quality assessment is proposed using the relation between SVD-Entropy and perceptual quality, based on a sequence of distorted images. Experiments show that the proposed method, compared with experimental psycho-visual scores, demonstrates a good consistency between these scores and stopping criterion measures that we obtain.

show abstract