Due to the loss of 3D information, accurate and robust 2D image feature matching remains challenging for many computer vision applications. This paper introduces a 2.5D feature that uses the disparity value from the light field Fourier disparity layer (FDL) as a rough proxy of scene depth. Without explicit depth estimation, a parameterized depth-degraded projection is proposed to construct the geometric transformation of paired features between two light fields. Then, we propose a parameterized learning solution to calculate the depth-degraded projection. This solution estimates a global constant fundamental matrix, a variable disparity-guided translation vector, and a depth compensation term using a very simple network. Although the 0.5D relative disparity provided by the FDL does not represent precise depth, it can also significantly reduce the depth ambiguity in feature matching. Therefore, the proposed solution achieves accurate feature-matching results by minimizing the sum of reprojection errors across all matching candidates. On the public light field feature-matching dataset, the proposed solution outperforms existing 2D image feature-matching solutions and light field feature-matching algorithms in terms of matching accuracy and robustness. The code is available online.

Deep-sea exploration has garnered increasing global interest. Non-uniform illumination and scattering from suspended particles in deep-sea environments significantly impair underwater image visibility. However, existing methods have limitations when dealing with strongly scattered scenes under non-uniform lighting conditions. To address these challenges, we introduce a novel deep-sea scattering model specifically tailored to the unique optical properties of the deep-sea environment, with a particular focus on the non-uniform active lighting conditions. We then propose a reconstruction algorithm that restores images by solving an optimization problem, integrating both the physical scattering model and a visual correction term. Evaluations across six deep-sea scenarios demonstrate superior performance, achieving 40% lower FADE scores and 5% higher HyperIQA scores than state-of-the-art methods, with notable improvements in texture preservation and color fidelity.

Deep generative models have proven to be effective priors for solving a variety of image processing problems. However, the learning of realistic image priors, based on a large number of parameters, requires a large amount of training data. It has been shown recently, with the so-called deep image prior (DIP), that randomly initialized neural networks can act as good image priors without learning. In this paper, we propose a deep generative model for light fields, which is compact and which does not require any training data other than the light field itself. To show the potential of the proposed generative model, we develop a complete light field compression scheme with quantization-aware learning and entropy coding of the quantized weights. Experimental results show that the proposed method yields very competitive results compared with state-of-the-art light field compression methods, both in terms of PSNR and MS-SSIM metrics.

In this paper, we present a learning-based framework for light field view synthesis from a subset of input views. Building upon a light-weight optical flow estimation network to obtain depth maps, our method employs two reconstruction modules in pixel and feature domains respectively. For the pixel-wise reconstruction, occlusions are explicitly handled by a disparity-dependent interpolation filter, whereas inpainting on disoccluded areas is learned by convolutional layers. Due to disparity inconsistencies, the pixel-based reconstruction may lead to blurriness in highly textured areas as well as on object contours. On the contrary, the feature-based reconstruction well performs on high frequencies, making the reconstruction in the two domains complementary. End-to-end learning is finally performed including a fusion module merging pixel and feature-based reconstructions. Experimental results show that our method achieves state-of-the-art performance on both synthetic and real-world datasets, moreover, it is even able to extend light fields' baseline by extrapolating high quality views without additional training.

We address the problem of light field dimensionality reduction for compression. We describe a local low rank approximation method using a parametric disparity model. The local support of the approximation is defined by super-rays. A super-ray can be seen as a set of super-pixels that are coherent across all light field views. A dedicated super-ray construction method is first described that constrains the super-pixels forming a given super-ray to be all of the same shape and size, dealing with occlusions. This constraint is needed so that the super-rays can be used as supports of angular dimensionality reduction based on low rank matrix approximation. The light field low rank assumption depends on how much the views are correlated, i.e. on how well they can be aligned by disparity compensation. We first introduce a parametric model describing the local variations of disparity within each super-ray. We then consider two methods for estimating the model parameters. The first method simply fits the model on an input disparity map. We then introduce a disparity estimation method using a low rank prior. This method alternatively searches for the best parameters of the disparity model and of the low rank approximation. We assess the proposed disparity parametric model, first assuming that the disparity is constant within a super-ray, and second by considering an affine disparity model. We show that using the proposed disparity parametric model and estimation algorithm gives an alignment of super-pixels across views that favours the low rank approximation compared with using disparity estimated with classical computer vision methods. The low rank matrix approximation is computed on the disparity compensated super-rays using a singular value decomposition (SVD). A coding algorithm is then described for the different components of the proposed disparity-compensated low rank approximation. Experimental results show performance gains, with a rate saving going up to 92.61%, compared with the JPEG Pleno anchor, for real light fields captured by a Lytro Illum camera. The rate saving goes up to 37.72% with synthetic light fields. The approach is also shown to outperform an HEVC-based light field compression scheme.

We describe a local low rank approximation method based on super-rays for light field compression. Super-rays can be seen as a set of super-pixels that are coherent across all light field views. A super-ray based disparity estimation method is proposed using a low rank prior, in order to be able to align all the super-pixels forming each super-ray. A dedicated super-ray construction method is described that constrains the super-pixels forming a given super-ray to be all of the same shape and size, dealing with occlusions. This constraint is needed so that the super-rays can be used as a support of angular dimensionality reduction based on low rank matrix approximation. A low rank matrix approximation is then computed on the disparity compensated super-rays using a singular value decomposition (SVD). A coding algorithm is then described for the different components of the resulting low rank approximation. Experimental results show performance gains compared with two reference light field coding schemes (HEVC-based scheme and JPEG-Pleno VM 1.1).

This paper proposes a learning based solution to disparity (depth) estimation for either densely or sparsely sampled light fields. Disparity between stereo pairs among a sparse subset of anchor views is first estimated by a fine-tuned FlowNet 2.0 network adapted to disparity prediction task. These coarse estimates are fused by exploiting the photo-consistency warping error, and refined by a Multi-view Stereo Refinement Network (MSRNet). The propagation of disparity from anchor viewpoints towards other viewpoints is performed by an occlusion-aware soft 3D reconstruction method. The experiments show that, both for dense and sparse light fields, our algorithm outperforms significantly the state-of-the-art algorithms, especially for subpixel accuracy.

In this paper, we propose a learning-based depth estimation framework suitable for both densely and sparsely sampled light fields. The proposed framework consists of three processing steps: initial depth estimation, fusion with occlusion handling, and refinement. The estimation can be performed from a flexible subset of input views. The fusion of initial disparity estimates, relying on two warping error measures, allows us to have an accurate estimation in occluded regions and along the contours. In contrast with methods relying on the computation of cost volumes, the proposed approach does not need any prior information on the disparity range. Experimental results show that the proposed method outperforms state-of-the-art light fields depth estimation methods, including prior methods based on deep neural architectures.

International audience

Building up on the advances in low rank matrix completion, this article presents a novel method for propagating the inpainting of the central view of a light field to all the other views. After generating a set of warped versions of the inpainted central view with random homographies, both the original light field views and the warped ones are vectorized and concatenated into a matrix. Because of the redundancy between the views, the matrix satisfies a low rank assumption enabling us to fill the region to inpaint with low rank matrix completion. To this end, a new matrix completion algorithm, better suited to the inpainting application than existing methods, is also developed in this paper. In its simple form, our method does not require any depth prior, unlike most existing light field inpainting algorithms. The method has then been extended to better handle the case where the area to inpaint contains depth discontinuities. In this case, a segmentation map of the different depth layers of the inpainted central view is required. This information is used to warp the depth layers with different homographies. Our experiments with natural light fields captured with plenoptic cameras demonstrate the robustness of the low rank approach to noisy data as well as large color and illumination variations between the views of the light field.

This paper studies the problem of low rank approximation of light fields for compression. A homography-based approximation method is proposed which jointly searches for homographies to align the different views of the light field together with the low rank approximation matrices. We first consider a global homography per view and show that depending on the variance of the disparity across views, the global homography is not sufficient to well-align the entire images. In a second step, we thus consider multiple homographies, one per region, the region being extracted using depth information. We first show the benefit of the joint optimization of the homographies together with the low-rank approximation. The resulting compact representation is then compressed using HEVC and the results are compared with those obtained by directly applying HEVC on the light field views re-structured as a video sequence. The experiments using different data sets show substantial PSNR-rate gain of our method, especially for real light fields.

The problem of finding nearest neighbours in terms of Euclidean distance, Hamming distance or other distance metric is a very common operation in computer vision and pattern recognition. In order to accelerate the search for the nearest neighbour in large collection datasets, many methods rely on the coarse-fine approach. In this paper we propose to combine Product Quantization (PQ) and binary neural associative memories to perform the coarse search. Our motivation lies in the fact that neural network dimensions of the representation associated with a set of k vectors is independent of k. We run experiments on TEXMEX SIFT1M and MNIST databases and observe significant improvements in terms of complexity of the search compared to raw PQ.

An extension to a recently introduced architecture of clique-based neural networks is presented. This extension makes it possible to store sequences with high efficiency. To obtain this property, network connections are provided with orientation and with flexible redundancy carried by both spatial and temporal redundancies, a mechanism of anticipation being introduced in the model. In addition to the sequence storage with high efficiency, this new scheme also offers biological plausibility. In order to achieve accurate sequence retrieval, a double-layered structure combining heteroassociation and autoassociation is also proposed.