Shadow and Specularity Priors for Intrinsic Light Field Decomposition

Alperovich, Anna; Johannsen, Ole; Strecke, Michael; Goldluecke, Bastian

doi:10.1007/978-3-319-78199-0_26

Cited by 3 publications

(7 citation statements)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our method outperforms recent light field based methods [34,2], and a single image deep network approach for intrinsic image decomposition [31]. Although we have only…”

Section: Resultsmentioning

confidence: 75%

“…The first one by Sulc et al [34] performs reflection separation based on specular flow. The second one is by Alperovich et al [2] and performs intrinsic light field decomposition. In addition, we compare to the network proposed by Shi et al [31], which uses a deep autoencoder for intrinsic images.…”

Section: Resultsmentioning

confidence: 99%

“…However, it only works for standard 2D images. To compare to the full decomposition [2], where the authors decompose the input light field into albedo, shading and specularity, we compute the diffuse component by multiplying albedo and Figure 9. Ablation study: Quantitative comparison of separation over nine previously unseen test datasets, and depth estimation for the two scenes from Figure 7.…”

Section: Resultsmentioning

confidence: 99%

“…Alperovich et al [2] Sulc et al [34] Shi et al [31] diffuse specular Figure 10. Comparison for a synthetic data set with two non-Lambertian objects with almost no texture, which is typically challenging for reflection separation.…”

Section: Oursmentioning

confidence: 99%

“…Comparison for a synthetic data set with two non-Lambertian objects with almost no texture, which is typically challenging for reflection separation. Both modeling approaches [2] and [34] fail to separate the specular component from the diffuse one. The CNNbased approach [31] successfully separates reflection components, but the diffuse one has some artifacts.…”

Section: Oursmentioning

confidence: 99%

See 4 more Smart Citations

Light Field Intrinsics with a Deep Encoder-Decoder Network

Alperovich

Johannsen

Strecke

et al. 2018

2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition

Self Cite

View full text Add to dashboard Cite

We present a fully convolutional autoencoder for light fields, which jointly encodes stacks of horizontal and vertical epipolar plane images through a deep network of residual layers. The complex structure of the light field is thus reduced to a comparatively low-dimensional representation, which can be decoded in a variety of ways. The different pathways of upconvolution we currently support are for disparity estimation and separation of the lightfield into diffuse and specular intrinsic components. The key idea is that we can jointly perform unsupervised training for the autoencoder path of the network, and supervised training for the other decoders. This way, we find features which are both tailored to the respective tasks and generalize well to datasets for which only example light fields are available. We provide an extensive evaluation on synthetic light field data, and show that the network yields good results on previously unseen real world data captured by a Lytro Illum camera and various gantries.

show abstract

“…Our method outperforms recent light field based methods [34,2], and a single image deep network approach for intrinsic image decomposition [31]. Although we have only…”

Section: Resultsmentioning

confidence: 75%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Oursmentioning

confidence: 99%

Section: Oursmentioning

confidence: 99%

See 3 more Smart Citations

Light Field Intrinsics with a Deep Encoder-Decoder Network

Alperovich

Johannsen

Strecke

et al. 2018

2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition

Self Cite

View full text Add to dashboard Cite

show abstract

Specular Highlight Removal for Real‐world Images

Zhang

Song

et al. 2019

Computer Graphics Forum

View full text Add to dashboard Cite

Removing specular highlight in an image is a fundamental research problem in computer vision and computer graphics. While various methods have been proposed, they typically do not work well for real‐world images due to the presence of rich textures, complex materials, hard shadows, occlusions and color illumination, etc. In this paper, we present a novel specular highlight removal method for real‐world images. Our approach is based on two observations of the real‐world images: (i) the specular highlight is often small in size and sparse in distribution; (ii) the remaining diffuse image can be represented by linear combination of a small number of basis colors with the sparse encoding coefficients. Based on the two observations, we design an optimization framework for simultaneously estimating the diffuse and specular highlight images from a single image. Specifically, we recover the diffuse components of those regions with specular highlight by encouraging the encoding coefficients sparseness using L0 norm. Moreover, the encoding coefficients and specular highlight are also subject to the non‐negativity according to the additive color mixing theory and the illumination definition, respectively. Extensive experiments have been performed on a variety of images to validate the effectiveness of the proposed method and its superiority over the previous methods.

show abstract