Sparse Set Of Images Research Articles

Fast radiance field reconstruction from sparse inputs

Neural Radiance Field (NeRF) has emerged as a powerful method in data-driven 3D reconstruction because of its simplicity and state-of-the-art performance. However, NeRF requires densely captured calibrated images and lengthy training and rendering time to realize high-resolution reconstruction. Thus, we propose a fast radiance field reconstruction method from a sparse set of images with silhouettes. Our approach integrates NeRF with Shape from Silhouette, a traditional 3D reconstruction method that uses silhouette information to fit the shape of an object. To combine NeRF’s implicit representation with Shape from Silhouette’s explicit representation, we propose a novel explicit–implicit radiance field representation consisting of voxel grids with confidence and feature embedding for geometry and a multilayer perceptron network to decode view-dependent color emission for appearance. We propose to make the reconstructed geometry compact by taking advantage of silhouette images, which can avoid the majority of artifacts in sparse input scenarios and speed up training and rendering. We also apply voxel dilating and pruning to refine the geometry prediction. In addition, we impose a total variation regularization on our model to encourage a smooth radiance field. Experiments on the DTU and the NeRF-Synthetic datasets show that our algorithm surpasses the existing baselines in terms of efficiency and accuracy.

Light Field View Synthesis via Aperture Disparity and Warping Confidence Map

This paper presents a learning-based approach to synthesize the view from an arbitrary camera position given a sparse set of images. A key challenge for this novel view synthesis arises from the reconstruction process, when the views from different input images may not be consistent due to obstruction in the light path. We overcome this by jointly modeling the epipolar property and occlusion in designing a convolutional neural network. We start by defining and computing the aperture disparity map, which approximates the parallax and measures the pixel-wise shift between two views. While this relates to free-space rendering and can fail near the object boundaries, we further develop a warping confidence map to address pixel occlusion in these challenging regions. The proposed method is evaluated on diverse real-world and synthetic light field scenes, and it shows better performance over several state-of-the-art techniques.

360 Panorama Synthesis from a Sparse Set of Images on a Low-Power Device

A full 360 $^\circ$ × 180 $^\circ$ image provides an unlimited field of view (FOV) and an immersive experience for the users without any loss of information of the surrounding. In this study, a deep learning based approach is proposed to synthesize a 360 $^\circ$ image from a sparse set of images captured with a limited FOV. The proposed network consists of a cascade of the FOV estimation network and the panorama synthesis network. We propose a hierarchical generative network to synthesize high quality 360 $^\circ$ panorama images. The design of progressive multi-scale generator and multiple discriminator reduces the high frequency artifact which is commonly observed in image synthesis using generative networks. The network is further compressed to run on a low-power device such as a smartphone using our proposed size and latency optimization. Experimental result demonstrates that the proposed method produces 360 $^\circ$ panorama with satisfactory image quality of up to 512 × 1024 resolution. It is also shown that the proposed method outperforms the alternative method and can be generalized for non-panoramic scenes and real images captured by a smartphone camera.

Scalable Coding of Plenoptic Images by Using a Sparse Set and Disparities.

One of the light field capturing techniques is the focused plenoptic capturing. By placing a microlens array in front of the photosensor, the focused plenoptic cameras capture both spatial and angular information of a scene in each microlens image and across microlens images. The capturing results in a significant amount of redundant information, and the captured image is usually of a large resolution. A coding scheme that removes the redundancy before coding can be of advantage for efficient compression, transmission, and rendering. In this paper, we propose a lossy coding scheme to efficiently represent plenoptic images. The format contains a sparse image set and its associated disparities. The reconstruction is performed by disparity-based interpolation and inpainting, and the reconstructed image is later employed as a prediction reference for the coding of the full plenoptic image. As an outcome of the representation, the proposed scheme inherits a scalable structure with three layers. The results show that plenoptic images are compressed efficiently with over 60 percent bit rate reduction compared with High Efficiency Video Coding intra coding, and with over 20 percent compared with an High Efficiency Video Coding block copying mode.

Image-based relighting from a sparse set of outdoor images

We propose a novel method for relighting the image of outdoor scenes viewed from a fixed camera based on a sparse set of images of the same scene under different illumination. Unlike previous methods which require capturing images under pre-designed lighting or employing the 3D model of the target objects, our method adopts the technique of basis image which encapsulates material and geometry information of the scene into one image. We present a new method to calculate the basis images of an outdoor scene based on the sampling images captured in different time of a day, the relighting images corresponding to new sunlight incidence directions with arbitrary intensity can then be generated with these basis images. Besides, the subordinate shadow effect adhering to the sun's movement is also simulated, producing a visually plausible relighting result. Experiments demonstrate the efficiency and validity of our approach.

Exploiting sparsity to accelerate noncontrast MR angiography in the context of parallel imaging

Noncontrast techniques for peripheral MR angiography are receiving renewed interest because of safety concerns about the use of gadolinium in patients with renal insufficiency. One class of techniques involves subtraction of dark-blood images acquired during fast systolic flow from bright-blood images obtained during slow diastolic flow. The goal of this work was to determine whether the inherent sparsity of the difference images could be exploited to achieve greater acceleration without loss of image quality in the context of generalized autocalibrating partially parallel acquisition (GRAPPA). It is shown that noise amplification at high acceleration factors can be reduced by performing subtraction on the raw data, before calculation of the GRAPPA weights, rather than on the final magnitude images. Use of the difference data to calculate the GRAPPA weights decreases the geometry factor (g-factor), because the difference data represent a sparse image set. This demonstrates an inherent property of GRAPPA and does not require the use of compressed sensing. Application of this approach to highly accelerated data from healthy volunteers resulted in similar depiction of large arteries to that obtained with low acceleration and standard reconstruction. However, visualization of very small vessels and arterial branches was compromised.

Illumination Recovery From Image With Cast Shadows Via Sparse Representation

In this paper, we propose using sparse representation for recovering the illumination of a scene from a single image with cast shadows, given the geometry of the scene. The images with cast shadows can be quite complex and, therefore, cannot be well approximated by low-dimensional linear subspaces. However, it can be shown that the set of images produced by a Lambertian scene with cast shadows can be efficiently represented by a sparse set of images generated by directional light sources. We first model an image with cast shadows composed of a diffusive part (without cast shadows) and a residual part that captures cast shadows. Then, we express the problem in an l(1)-regularized least-squares formulation, with nonnegativity constraints (as light has to be non-negative at any point in space). This sparse representation enjoys an effective and fast solution thanks to recent advances in compressive sensing. In experiments on synthetic and real data, our approach performs favorably in comparison with several previously proposed methods.

Modeling Complex Unfoliaged Trees from a Sparse Set of Images

AbstractWe present a novel image‐based technique for modeling complex unfoliaged trees. Existing tree modeling tools either require capturing a large number of views for dense 3D reconstruction or rely on user inputs and botanic rules to synthesize natural‐looking tree geometry. In this paper, we focus on faithfully recovering real instead of realistically‐looking tree geometry from a sparse set of images. Our solution directly integrates 2D/3D tree topology as shape priors into the modeling process. For each input view, we first estimate a 2D skeleton graph from its matte image and then find a 2D skeleton tree from the graph by imposing tree topology. We develop a simple but effective technique for computing the optimal 3D skeleton tree most consistent with the 2D skeletons. For each edge in the 3D skeleton tree, we further apply volumetric reconstruction to recover its corresponding curved branch. Finally, we use piecewise cylinders to approximate each branch from the volumetric results. We demonstrate our framework on a variety of trees to illustrate the robustness and usefulness of our technique.

Reflectance sharing: predicting appearance from a sparse set of images of a known shape

Three-dimensional appearance models consisting of spatially varying reflectance functions defined on a known shape can be used in analysis-by-synthesis approaches to a number of visual tasks. The construction of these models requires the measurement of reflectance, and the problem of recovering spatially varying reflectance from images of known shape has drawn considerable interest. To date, existing methods rely on either: (1) low-dimensional (e.g., parametric) reflectance models, or (2) large data sets involving thousands of images (or more) per object. Appearance models based on the former have limited accuracy and generality since they require the selection of a specific reflectance model a priori, and while approaches based on the latter may be suitable for certain applications, they are generally too costly and cumbersome to be used for image analysis. We present an alternative approach that seeks to combine the benefits of existing methods by enabling the estimation of a nonparametric spatially varying reflectance function from a small number of images. We frame the problem as scattered-data interpolation in a mixed spatial and angular domain, and we present a theory demonstrating that the angular accuracy of a recovered reflectance function can be increased in exchange for a decrease in its spatial resolution. We also present a practical solution to this interpolation problem using a new representation of reflectance based on radial basis functions. This representation is evaluated experimentally by testing its ability to predict appearance under novel view and lighting conditions. Our results suggest that since reflectance typically varies slowly from point to point over much of an object's surface, we can often obtain a nonparametric reflectance function from a sparse set of images. In fact, in some cases, we can obtain reasonable results in the limiting case of only a single input image.

Photogrammetric reconstruction of free-form objects with curvilinear structures

The shapes of many natural or man-made objects have curve features. The images of such curves usually do not have sufficient distinctive features to apply conventional feature-based reconstruction algorithms. In this paper, we introduce a photogrammetric method for recovering free-form objects with curvilinear structures. Our method chooses to obtain the topology and geometry of a sparse 3D wireframe of the object first instead of directly recovering a surface or volume model. Surface patches covering the object are then constructed to interpolate the curves in this wireframe while satisfying certain heuristics such as minimal bending energy. The result is an object surface model with curvilinear structures from a sparse set of images. We can produce realistic texture-mapped renderings of the object model from arbitrary viewpoints. Reconstruction results on multiple real objects are presented to demonstrate the effectiveness of our approach.